# Vercel AI Gateway Supported Providers and Models

> List of all 431 models available through the [Vercel AI Gateway](https://vercel.com/ai-gateway).
> The [model list API](https://ai-gateway.vercel.sh/v1/models) returns 499 model IDs as JSON, including separate IDs for batch, fast, and free tiers. No API key required.
> For API documentation, see https://vercel.com/docs/ai-gateway

| Model | Type | Context | Input | Output | Providers | Tags |
|-------|------|---------|-------|--------|-----------|------|
| alibaba/qwen-3-32b | chat | 128K | $0.08/1M | $0.28/1M | bedrock, deepinfra | reasoning, tool-use, structured-output, open-weights |
| alibaba/qwen3-coder-30b-a3b | chat | 262.1K | $0.15/1M | $0.60/1M | bedrock | tool-use, structured-output, open-weights |
| alibaba/qwen3-max-thinking | chat | 256K | $1.20/1M | $6.00/1M | alibaba | reasoning, tool-use, structured-output |
| alibaba/qwen3-vl-235b-a22b-instruct | chat | 262.1K | $0.20/1M | $0.88/1M | deepinfra | tool-use, vision, structured-output, open-weights |
| alibaba/qwen3.5-flash | chat | 1M | $0.10/1M | $0.40/1M | alibaba | vision, file-input, reasoning, tool-use, explicit-caching |
| alibaba/qwen3.5-plus | chat | 1M | $0.40/1M | $2.40/1M | alibaba | vision, file-input, reasoning, tool-use, explicit-caching |
| alibaba/qwen3.6-27b | chat | 256K | $0.60/1M | $3.60/1M | alibaba | reasoning, tool-use, file-input, vision, open-weights |
| alibaba/qwen3.6-plus | chat | 1M | $0.50/1M | $3.00/1M | alibaba, togetherai | reasoning, tool-use, vision, file-input, explicit-caching |
| alibaba/qwen3.7-flash | chat | 991K | $0.03/1M | $0.13/1M | alibaba | reasoning, tool-use, implicit-caching, file-input, vision, explicit-caching, structured-output |
| alibaba/qwen3.7-max | chat | 991K | $2.50/1M | $7.50/1M | alibaba | implicit-caching, reasoning, tool-use, explicit-caching, structured-output |
| alibaba/qwen3.7-plus | chat | 1M | $0.40/1M | $1.60/1M | alibaba | reasoning, tool-use, implicit-caching, file-input, vision, explicit-caching |
| alibaba/qwen3.8-27b | chat | 1M | $0.06/1M | $0.32/1M | cerebras, runinfra, deepinfra, alibaba, parasail, novita, wafer | implicit-caching, reasoning, tool-use, vision, structured-output, open-weights |
| alibaba/qwen3.8-flash | chat | 991K | $0.15/1M | $0.47/1M | alibaba | reasoning, tool-use, implicit-caching, file-input, vision, explicit-caching, structured-output |
| alibaba/qwen3.8-max | chat | 1M | $2.00/1M | $6.00/1M | alibaba, fireworks | reasoning, tool-use, implicit-caching, vision, explicit-caching, structured-output |
| alibaba/qwen3.8-max-prime | chat | 1M | $4.00/1M | $12.00/1M | alibaba | implicit-caching, reasoning, structured-output, tool-use, vision, explicit-caching |
| alibaba/qwen3.8-omni-flash | chat | 1M | $0.15/1M | $0.47/1M | alibaba | implicit-caching, reasoning, structured-output, tool-use, vision |
| alibaba/qwen-3-235b | chat | 262.1K | $0.09/1M | $0.55/1M | deepinfra, vertex | tool-use, structured-output, open-weights |
| alibaba/qwen3-coder | chat | 262.1K | $0.22/1M | $1.80/1M | vertex | implicit-caching, tool-use, structured-output, open-weights |
| alibaba/qwen3-coder-next | chat | 256K | $0.50/1M | $1.20/1M | bedrock | tool-use, structured-output, open-weights |
| alibaba/qwen3-embedding-0.6b | embedding | 32.8K | $0.01/1M | $0.00/1M | deepinfra | open-weights |
| alibaba/qwen3-embedding-4b | embedding | 32.8K | $0.02/1M | $0.00/1M | deepinfra | open-weights |
| alibaba/qwen3-embedding-8b | embedding | 41K | $0.01/1M | $0.00/1M | deepinfra, nebius | open-weights |
| alibaba/qwen3-next-80b-a3b-instruct | chat | 262.1K | $0.15/1M | $1.20/1M | vertex, deepinfra | tool-use, structured-output, open-weights |
| alibaba/qwen3-next-80b-a3b-thinking | chat | 262.1K | $0.15/1M | $1.20/1M | vertex | reasoning, tool-use, structured-output, open-weights |
| alibaba/qwen3-vl-instruct | chat | 262.1K | $0.20/1M | $0.88/1M | deepinfra | implicit-caching, tool-use, vision, open-weights |
| alibaba/qwen-3-14b | chat | 41.0K | $0.12/1M | $0.24/1M | deepinfra | reasoning, tool-use, structured-output, open-weights |
| alibaba/qwen-3-30b | chat | 41.0K | $0.12/1M | $0.50/1M | deepinfra | reasoning, tool-use, structured-output, open-weights |
| alibaba/qwen3.8-2.4t-a95b | chat | 262.1K | $2.00/1M | $6.00/1M | deepinfra, fireworks, gmicloud, novita, modal, togetherai | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| alibaba/qwen3.8-max-0902 | chat | 991K | $2.00/1M | $6.00/1M | alibaba | reasoning, tool-use, vision, file-input, implicit-caching, explicit-caching, structured-output |
| alibaba/wan-v2.5-t2v-preview | video | 0 | $0.00/1M | $0.00/1M | alibaba | text-to-video |
| alibaba/wan-v2.6-i2v | video | 0 | $0.00/1M | $0.00/1M | alibaba | image-to-video |
| alibaba/wan-v2.6-i2v-flash | video | 0 | $0.00/1M | $0.00/1M | alibaba | image-to-video |
| alibaba/wan-v2.6-r2v | video | 0 | $0.00/1M | $0.00/1M | alibaba | reference-to-video |
| alibaba/wan-v2.6-r2v-flash | video | 0 | $0.00/1M | $0.00/1M | alibaba | reference-to-video |
| alibaba/wan-v2.6-t2v | video | 0 | $0.00/1M | $0.00/1M | alibaba | text-to-video |
| alibaba/wan-v2.7-r2v | video | 0 | $0.00/1M | $0.00/1M | alibaba | reference-to-video |
| alibaba/wan-v2.7-t2v | video | 0 | $0.00/1M | $0.00/1M | alibaba | text-to-video |
| alibaba/wan-v3.0-video | video | 0 | $0.00/1M | $0.00/1M | alibaba | text-to-video, image-to-video, reference-to-video, audio-generation, vision, video-input, audio-input |
| alibaba/wan-v3.0-video-prime | video | 0 | $0.00/1M | $0.00/1M | alibaba | text-to-video, image-to-video, reference-to-video, audio-generation, vision, video-input, audio-input |
| amazon/nova-2-lite | chat | 1M | $0.33/1M | $2.75/1M | bedrock | file-input, reasoning, tool-use, vision, structured-output |
| amazon/nova-lite | chat | 300K | $0.06/1M | $0.24/1M | bedrock | file-input, tool-use, vision, structured-output |
| amazon/nova-micro | chat | 128K | $0.04/1M | $0.14/1M | bedrock | tool-use, structured-output |
| amazon/nova-pro | chat | 300K | $0.80/1M | $3.20/1M | bedrock | file-input, tool-use, vision, structured-output |
| amazon/titan-embed-text-v2 | embedding | 0 | $0.02/1M | $0.00/1M | bedrock |  |
| anthropic/claude-fable-5 | chat | 1M | $10.00/1M | $50.00/1M | anthropic, bedrock, vertexAnthropic | reasoning, tool-use, explicit-caching, file-input, vision, structured-output |
| anthropic/claude-fable-5.1 | chat | 1M | $10.00/1M | $50.00/1M | anthropic, bedrock, vertexAnthropic | reasoning, tool-use, explicit-caching, file-input, vision, structured-output |
| anthropic/claude-haiku-4.5 | chat | 200K | $1.00/1M | $5.00/1M | anthropic, bedrock, vertexAnthropic, claudeaws | explicit-caching, file-input, reasoning, tool-use, vision, web-search, structured-output |
| anthropic/claude-haiku-5.5 | chat | 1M | $0.10/1M | $0.50/1M | anthropic, vertexAnthropic, bedrock, claudeaws | explicit-caching, file-input, reasoning, tool-use, vision, web-search, structured-output |
| anthropic/claude-opus-4.5 | chat | 200K | $5.00/1M | $25.00/1M | anthropic, bedrock, vertexAnthropic | explicit-caching, file-input, reasoning, tool-use, vision, web-search, structured-output |
| anthropic/claude-opus-4.6 | chat | 1M | $5.00/1M | $25.00/1M | anthropic, bedrock, vertexAnthropic | tool-use, reasoning, vision, file-input, explicit-caching, web-search, structured-output |
| anthropic/claude-opus-4.7 | chat | 1M | $5.00/1M | $25.00/1M | anthropic, bedrock, vertexAnthropic | tool-use, reasoning, vision, file-input, explicit-caching, web-search, structured-output |
| anthropic/claude-opus-4.8 | chat | 1M | $5.00/1M | $25.00/1M | anthropic, bedrock, vertexAnthropic, claudeaws | tool-use, reasoning, vision, file-input, explicit-caching, web-search, fast, structured-output |
| anthropic/claude-opus-5 | chat | 1M | $5.00/1M | $25.00/1M | anthropic, bedrock, vertexAnthropic, claudeaws | tool-use, reasoning, vision, file-input, explicit-caching, web-search, fast, structured-output |
| anthropic/claude-opus-5.5 | chat | 1M | $4.00/1M | $20.00/1M | anthropic, vertexAnthropic, claudeaws, bedrock | explicit-caching, fast, file-input, reasoning, tool-use, vision, web-search, structured-output |
| anthropic/claude-sonnet-4.5 | chat | 1M | $3.00/1M | $15.00/1M | anthropic, bedrock, vertexAnthropic | explicit-caching, file-input, reasoning, tool-use, vision, web-search, structured-output |
| anthropic/claude-sonnet-4.6 | chat | 1M | $3.00/1M | $15.00/1M | anthropic, vertexAnthropic, bedrock, claudeaws | file-input, reasoning, tool-use, vision, explicit-caching, web-search, structured-output |
| anthropic/claude-sonnet-5 | chat | 1M | $2.00/1M | $10.00/1M | anthropic, vertexAnthropic, bedrock, claudeaws | reasoning, tool-use, explicit-caching, file-input, vision, web-search, structured-output |
| anthropic/claude-sonnet-5.5 | chat | 1M | $2.00/1M | $10.00/1M | anthropic, vertexAnthropic, bedrock, claudeaws | explicit-caching, file-input, reasoning, tool-use, vision, web-search, structured-output |
| arcee-ai/trinity-large-thinking | chat | 262.1K | $0.25/1M | $0.80/1M | arcee-ai | reasoning, tool-use, structured-output, open-weights |
| baseten/elephant-1 | evaluation | 16.4K | $0.04/1M | $0.00/1M | baseten | open-weights |
| baseten/mammoth-1 | evaluation | 16.4K | $0.10/1M | $0.00/1M | baseten | open-weights |
| bfl/flux-3-video | video | 0 | $0.00/1M | $0.00/1M | bfl | video-generation, text-to-video, image-to-video, video-editing |
| bfl/flux-3-image | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation, image-editing |
| bfl/flux-3-image-klein | image | 0 | $0.00/1M | $0.00/1M | bfl | image-editing, image-generation, open-weights |
| bfl/flux-pro-1.0-fill | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-kontext-max | image | 512 | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-kontext-pro | image | 512 | $0.00/1M | $0.00/1M | bfl, prodia | image-generation |
| bfl/flux-2-flex | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-2-klein-4b | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation, open-weights |
| bfl/flux-2-klein-9b | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation, open-weights |
| bfl/flux-2-max | image | 67.3K | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-2-pro | image | 67.3K | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-pro-1.1 | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation |
| bfl/flux-pro-1.1-ultra | image | 0 | $0.00/1M | $0.00/1M | bfl | image-generation |
| bytedance/seed-1.8 | chat | 256K | $0.25/1M | $2.00/1M | bytedance | reasoning, tool-use, vision |
| bytedance/seed-1.6 | chat | 256K | $0.25/1M | $2.00/1M | bytedance | reasoning, tool-use, vision |
| bytedance/seed-2.1-turbo | chat | 262.1K | $0.50/1M | $2.50/1M | bytedance | reasoning, tool-use, video-input, vision |
| bytedance/seedance-2.0 | video | 0 | $0.00/1M | $0.00/1M | bytedance | video-generation, vision, image-to-video, text-to-video, reference-to-video |
| bytedance/seedance-2.0-fast | video | 0 | $0.00/1M | $0.00/1M | bytedance | vision, video-generation, text-to-video, image-to-video, reference-to-video |
| bytedance/seedance-2.0-mini | video | 0 | $0.00/1M | $0.00/1M | bytedance | vision, video-generation, text-to-video, image-to-video, reference-to-video |
| bytedance/seedance-2.5 | video | 0 | $0.00/1M | $0.00/1M | bytedance | video-generation |
| bytedance/seedance-v1.0-pro | video | 0 | $0.00/1M | $0.00/1M | bytedance | video-generation, text-to-video |
| bytedance/seedance-v1.0-pro-fast | video | 0 | $0.00/1M | $0.00/1M | bytedance | video-generation, text-to-video |
| bytedance/seedance-v1.5-pro | video | 0 | $0.00/1M | $0.00/1M | bytedance | video-generation, text-to-video |
| bytedance/seedream-4.0 | image | 0 | $0.00/1M | $0.00/1M | bytedance | image-generation |
| bytedance/seedream-4.5 | image | 0 | $0.00/1M | $0.00/1M | bytedance | image-generation |
| bytedance/seedream-5.0-lite | image | 0 | $0.00/1M | $0.00/1M | bytedance | image-generation |
| bytedance/seedream-5.0-pro | image | 0 | $0.00/1M | $0.00/1M | bytedance | image-generation |
| callstack/apex | chat | 262.1K | $0.50/1M | $3.00/1M | callstack | implicit-caching, reasoning, vision, tool-use |
| cohere/rerank-v3.5 | reranking | 4.1K | $0.00/1M | $0.00/1M | bedrock | reranking |
| cohere/rerank-v4-fast | reranking | 32K | $0.00/1M | $0.00/1M | cohere | reranking |
| cohere/rerank-v4-pro | reranking | 32K | $0.00/1M | $0.00/1M | cohere | reranking |
| cohere/rerank-v5-fast | reranking | 32.8K | $0.02/1M | $0.00/1M | cohere | reranking |
| cohere/rerank-v5-pro | reranking | 32.8K | $0.05/1M | $0.00/1M | cohere | reranking |
| cohere/command-a | chat | 256K | $2.50/1M | $10.00/1M | cohere | tool-use, structured-output, open-weights |
| cohere/embed-v5.0-fast | embedding | 128K | $0.08/1M | $0.00/1M | cohere |  |
| cohere/embed-v5.0-pro | embedding | 128K | $0.12/1M | $0.00/1M | cohere |  |
| cohere/embed-v4.0 | embedding | 128K | $0.12/1M | $0.00/1M | cohere, bedrock |  |
| convaiinnovations/laya | evaluation | 8.2K | $0.00/1M | $0.00/1M | boundless | free |
| deepseek/deepseek-r1 | chat | 160K | $1.35/1M | $5.40/1M | deepinfra, bedrock | implicit-caching, reasoning, tool-use, open-weights |
| deepseek/deepseek-v3.1 | chat | 163.8K | $0.25/1M | $0.95/1M | deepinfra, sambanova | implicit-caching, reasoning, tool-use, structured-output, open-weights |
| deepseek/deepseek-v3.2 | chat | 163.8K | $0.26/1M | $0.38/1M | deepinfra, bedrock | implicit-caching, tool-use, open-weights |
| deepseek/deepseek-v3.2-thinking | chat | 163.8K | $0.26/1M | $0.38/1M | bedrock, deepinfra | reasoning, tool-use, structured-output, open-weights |
| deepseek/deepseek-v4-flash | chat | 1.0M | $0.06/1M | $0.18/1M | digitalocean, novita, deepinfra, azure, alibaba, baseten, wafer, gmicloud | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| deepseek/deepseek-v4-flash-0731 | chat | 1.0M | $0.04/1M | $0.15/1M | baseten, deepinfra, novita, alibaba, wafer, runware, morph, parasail, digitalocean, streamlake, inceptron, relace, boundless, gmicloud | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| deepseek/deepseek-v4-flash-vision-exp | chat | 1.0M | $0.22/1M | $0.65/1M | novita, deepinfra | implicit-caching, reasoning, tool-use, vision, open-weights |
| deepseek/deepseek-v4-pro | chat | 1.0M | $0.66/1M | $1.98/1M | deepseek, novita, deepinfra, baseten, azure, alibaba | reasoning, tool-use, implicit-caching, open-weights |
| deepseek/deepseek-v4-pro-0813 | chat | 1.0M | $0.66/1M | $1.98/1M | deepseek, baseten, gmicloud, alibaba, parasail, digitalocean, streamlake, deepinfra, novita, tencent | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| deepseek/deepseek-v4.1-flash | chat | 1.0M | $0.04/1M | $0.30/1M | deepseek, novita, deepinfra, morph, fireworks, gmicloud, parasail, baseten, wafer, modal, particle, togetherai, alibaba, relace, boundless, runware, inference-net, runinfra, vertex, pareto, digitalocean, tencent | reasoning, tool-use, implicit-caching, vision, structured-output, open-weights |
| deepseek/deepseek-v4.1-flash-1017 | chat | 1.0M | $0.04/1M | $0.29/1M | deepseek, deepinfra, inference-net, morph, runinfra, parasail, baseten, relace, pareto, fireworks, alibaba, wafer, novita, togetherai, particle, gmicloud, runware, boundless, digitalocean, streamlake | reasoning, tool-use, implicit-caching, file-input, vision, open-weights |
| deepseek/deepseek-v4.1-flash-1017-fast | chat | 1.0M | $0.60/1M | $2.40/1M | baseten | implicit-caching, reasoning, structured-output, tool-use, vision, open-weights |
| deepseek/deepseek-v4.1-flash-fast | chat | 1.0M | $0.60/1M | $2.40/1M | baseten | reasoning, tool-use, implicit-caching, vision, structured-output, open-weights |
| elevenlabs/eleven-flash-v2 | speech | 0 | $20.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/eleven-flash-v2.5 | speech | 0 | $20.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/eleven-multilingual-v2 | speech | 0 | $40.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/music-v2 | speech | 0 | $0.00/1M | $0.00/1M | elevenlabs | music |
| elevenlabs/music-v2.5 | speech | 0 | $0.00/1M | $0.00/1M | elevenlabs | music |
| elevenlabs/eleven-realtime-v1-max | realtime | 0 | $0.90/1M | $4.50/1M | elevenlabs |  |
| elevenlabs/eleven-realtime-v1-mini | realtime | 0 | $0.10/1M | $3.00/1M | elevenlabs |  |
| elevenlabs/sound-effects-v2 | speech | 0 | $0.00/1M | $0.00/1M | elevenlabs | sound-effects |
| elevenlabs/eleven-v3 | speech | 0 | $40.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/eleven-v3-conversational | speech | 0 | $20.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/eleven-v4 | speech | 0 | $40.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/eleven-v4-turbo | speech | 0 | $20.00/1M | $0.00/1M | elevenlabs | websocket-speech |
| elevenlabs/scribe-v2 | transcription | 0 | $0.00/1M | $0.00/1M | elevenlabs |  |
| elevenlabs/scribe-v2-realtime | transcription | 0 | $0.00/1M | $0.00/1M | elevenlabs | websockets |
| elorian/glimpse | chat | 220K | $0.60/1M | $3.00/1M | elorian | implicit-caching, reasoning, video-input, vision |
| fireworks/ember-1 | chat | 1.0M | $3.00/1M | $15.00/1M | fireworks | implicit-caching, reasoning, tool-use, vision |
| fish-audio/drama-3 | speech | 0 | $10.00/1M | $0.00/1M | fish-audio |  |
| fish-audio/s1 | speech | 0 | $15.00/1M | $0.00/1M | fish-audio |  |
| fish-audio/s2-pro | speech | 0 | $15.00/1M | $0.00/1M | fish-audio |  |
| fish-audio/s2.1-pro | speech | 0 | $15.00/1M | $0.00/1M | fish-audio |  |
| fish-audio/transcribe-1 | transcription | 0 | $0.00/1M | $0.00/1M | fish-audio |  |
| google/gemini-2.5-flash | chat | 1M | $0.30/1M | $2.50/1M | vertex, google | file-input, reasoning, tool-use, vision, web-search, implicit-caching, structured-output |
| google/gemini-2.5-flash-lite | chat | 1.0M | $0.10/1M | $0.40/1M | vertex, google | file-input, reasoning, tool-use, vision, web-search, implicit-caching, structured-output |
| google/gemini-2.5-pro | chat | 1.0M | $1.25/1M | $10.00/1M | vertex, google | file-input, reasoning, tool-use, vision, web-search, tiered-cost, implicit-caching, structured-output |
| google/gemini-3-flash | chat | 1M | $0.50/1M | $3.00/1M | vertex, google | reasoning, tool-use, file-input, vision, web-search, tiered-cost, implicit-caching, structured-output |
| google/gemini-3.1-flash-image | chat | 131.1K | $0.50/1M | $3.00/1M | vertex, google | image-generation, reasoning, vision, web-search, structured-output |
| google/gemini-3.1-flash-image-preview | chat | 131.1K | $0.50/1M | $3.00/1M | google | image-generation, implicit-caching, reasoning, vision, web-search, structured-output |
| google/gemini-3.1-flash-lite | chat | 1M | $0.25/1M | $1.50/1M | google, vertex | reasoning, tool-use, implicit-caching, file-input, vision, web-search, structured-output |
| google/gemini-3.1-flash-lite-image | chat | 65.5K | $0.25/1M | $1.50/1M | google, vertex | image-generation, implicit-caching, reasoning, vision, web-search |
| google/gemini-3.1-pro-preview | chat | 1M | $2.00/1M | $12.00/1M | google, vertex | file-input, tool-use, reasoning, vision, web-search, tiered-cost, implicit-caching, structured-output |
| google/gemini-3.5-flash | chat | 1M | $1.50/1M | $9.00/1M | google, vertex | reasoning, file-input, vision, tool-use, web-search, implicit-caching, video-input, structured-output |
| google/gemini-3.5-flash-lite | chat | 1M | $0.30/1M | $2.50/1M | google, vertex | reasoning, tool-use, implicit-caching, file-input, vision, web-search, video-input, structured-output |
| google/gemini-3.5-transcribe | transcription | 0 | $2.00/1M | $12.00/1M | google |  |
| google/gemini-3.5-transcribe-live | transcription | 0 | $0.00/1M | $0.00/1M | google | websockets |
| google/gemini-3.6-flash | chat | 1M | $0.75/1M | $3.75/1M | vertex, google | reasoning, file-input, tool-use, vision, implicit-caching, web-search, video-input, structured-output |
| google/gemini-3.7-flash | chat | 1M | $0.75/1M | $3.75/1M | google, vertex | reasoning, file-input, vision, tool-use, web-search, implicit-caching, video-input, structured-output |
| google/gemini-3.8-flash | chat | 1M | $0.75/1M | $3.75/1M | google, vertex | reasoning, file-input, vision, tool-use, web-search, implicit-caching, video-input, structured-output |
| google/gemini-3.8-flash-tts | speech | 8.2K | $0.50/1M | $9.00/1M | google, vertex |  |
| google/gemini-3.8-flash-lite-tts | speech | 8.2K | $0.50/1M | $6.00/1M | google, vertex |  |
| google/gemini-3.8-live | realtime | 0 | $0.75/1M | $4.50/1M | google | websockets |
| google/gemini-3.8-live-extended-thinking | realtime | 0 | $0.75/1M | $4.50/1M | google | websockets |
| google/gemini-4-argon | chat | 1M | $2.00/1M | $10.00/1M | google, vertex | reasoning, file-input, vision, tool-use, web-search, implicit-caching, video-input, structured-output |
| google/gemini-embedding-001 | embedding | 0 | $0.15/1M | $0.00/1M | google, vertex |  |
| google/gemini-embedding-2 | embedding | 0 | $0.20/1M | $0.00/1M | google |  |
| google/gemini-nano-banana-2.1 | chat | 131.1K | $1.50/1M | $7.50/1M | google, vertex | image-generation, implicit-caching, reasoning, vision, web-search, structured-output |
| google/gemini-omni-flash-preview | chat | 1M | $1.50/1M | $9.00/1M | vertex, google | file-input, reasoning, vision, video-generation, video-input, structured-output |
| google/gemma-4-26b-a4b-it | chat | 1.0M | $0.13/1M | $0.40/1M | novita, parasail, vertex, gmicloud | file-input, reasoning, tool-use, vision, structured-output, open-weights |
| google/gemma-4-31b-it | chat | 1.0M | $0.14/1M | $0.40/1M | novita, parasail, friendli, gmicloud | file-input, reasoning, tool-use, vision, structured-output, open-weights |
| google/gemini-2.5-flash-image | chat | 32.8K | $0.30/1M | $2.50/1M | google, vertex | image-generation, web-search, vision |
| google/gemini-3-pro-image | chat | 65.5K | $2.00/1M | $12.00/1M | google, vertex | image-generation, web-search, implicit-caching, vision, structured-output |
| google/text-embedding-005 | embedding | 0 | $0.03/1M | $0.00/1M | vertex |  |
| google/text-multilingual-embedding-002 | embedding | 0 | $0.03/1M | $0.00/1M | vertex |  |
| google/veo-3.0-generate-001 | video | 0 | $0.00/1M | $0.00/1M | vertex | text-to-video, image-to-video |
| google/veo-3.0-fast-generate-001 | video | 0 | $0.00/1M | $0.00/1M | vertex | video-generation |
| google/veo-3.1-generate-001 | video | 0 | $0.00/1M | $0.00/1M | vertex | text-to-video, image-to-video |
| google/veo-3.1-fast-generate-001 | video | 0 | $0.00/1M | $0.00/1M | vertex | text-to-video, image-to-video |
| google/veo-3.1-lite-generate-001 | video | 0 | $0.00/1M | $0.00/1M | vertex | text-to-video |
| heygen/heygen-video-1 | video | 0 | $0.00/1M | $0.00/1M | heygen | text-to-video, image-to-video, vision, video-generation |
| inception/mercury-2 | chat | 128K | $0.25/1M | $0.75/1M | inception | reasoning, tool-use, structured-output |
| inception/mercury-2.5 | chat | 260K | $0.04/1M | $0.15/1M | inception | implicit-caching, reasoning, tool-use, structured-output |
| inception/mercury-3 | chat | 1M | $0.10/1M | $0.35/1M | inception | implicit-caching, reasoning, structured-output, tool-use |
| inception/mercury-coder-small | chat | 32K | $0.25/1M | $1.00/1M | inception | tool-use, structured-output |
| inclusionai/ling-3.0-flash | chat | 256K | $0.02/1M | $0.06/1M | novita | reasoning, tool-use, implicit-caching, open-weights |
| inclusionai/ling-3.0-flash-fin | chat | 262.1K | $0.06/1M | $0.18/1M | novita, deepinfra | reasoning, tool-use, open-weights |
| inclusionai/ling-3.0-flash-sante | chat | 256K | $0.07/1M | $0.22/1M | novita | reasoning, tool-use, implicit-caching, open-weights |
| inclusionai/ling-3.0-flash-vl | chat | 256K | $0.07/1M | $0.22/1M | novita | reasoning, tool-use, video-input, vision, open-weights |
| inclusionai/ling-3.1-flash | chat | 262.1K | $0.30/1M | $0.90/1M | novita | reasoning, tool-use, implicit-caching, open-weights |
| inference-net/schematron-v2-small | chat | 128K | $0.05/1M | $0.23/1M | inference-net | structured-output |
| inference-net/schematron-v2-turbo | chat | 128K | $0.03/1M | $0.15/1M | inference-net | structured-output |
| interfaze/interfaze-beta | chat | 1M | $1.50/1M | $3.50/1M | interfaze | file-input, reasoning, tool-use, vision, structured-output |
| klingai/kling-v2.5-turbo-i2v | video | 0 | $0.00/1M | $0.00/1M | klingai | image-to-video, audio-generation |
| klingai/kling-v2.5-turbo-t2v | video | 0 | $0.00/1M | $0.00/1M | klingai | text-to-video, audio-generation |
| klingai/kling-v2.6-i2v | video | 0 | $0.00/1M | $0.00/1M | klingai | image-to-video, audio-generation |
| klingai/kling-v2.6-motion-control | video | 0 | $0.00/1M | $0.00/1M | klingai | motion-control |
| klingai/kling-v2.6-t2v | video | 0 | $0.00/1M | $0.00/1M | klingai | text-to-video, audio-generation |
| klingai/kling-v3.0-i2v | video | 0 | $0.00/1M | $0.00/1M | klingai | image-to-video, multi-shot, audio-generation |
| klingai/kling-v3.0-motion-control | video | 0 | $0.00/1M | $0.00/1M | klingai | video-generation, motion-control |
| klingai/kling-v3.0-t2v | video | 0 | $0.00/1M | $0.00/1M | klingai | text-to-video, multi-shot, audio-generation |
| liquid/lfm2.5-24b-a2b | chat | 128K | $0.05/1M | $0.70/1M | liquid | implicit-caching, reasoning, structured-output, tool-use |
| liquid/d1 | evaluation | 65.5K | $0.02/1M | $0.00/1M | liquid | vision |
| liquid/d1-ultra | evaluation | 32.8K | $0.04/1M | $0.00/1M | liquid | vision |
| meituan/longcat-2.5-preview | chat | 1.0M | $0.30/1M | $1.20/1M | meituan | implicit-caching, reasoning, tool-use, vision, structured-output |
| meta/llama-3.1-70b | chat | 128K | $0.72/1M | $0.72/1M | bedrock | tool-use, open-weights |
| meta/llama-3.1-8b | chat | 128K | $0.02/1M | $0.05/1M | bedrock, novita | tool-use, open-weights |
| meta/llama-3.3-70b | chat | 128K | $0.72/1M | $0.72/1M | bedrock | tool-use, open-weights |
| meta/llama-4-maverick | chat | 128K | $0.24/1M | $0.97/1M | bedrock | tool-use, vision, open-weights |
| meta/llama-4-scout | chat | 131.1K | $0.17/1M | $0.66/1M | deepinfra, bedrock | tool-use, vision, structured-output, open-weights |
| meta/muse-glimmer-30b | chat | 131.1K | $0.35/1M | $1.50/1M | togetherai | vision, implicit-caching, reasoning, tool-use, structured-output |
| meta/muse-image-1.0 | image | 0 | $0.00/1M | $0.00/1M | meta | image-generation, vision |
| meta/muse-spark-1.1 | chat | 1.0M | $1.25/1M | $4.25/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| meta/muse-spark-1.2 | chat | 1.0M | $1.25/1M | $4.25/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| meta/muse-spark-1.2-contributor | chat | 1.0M | $0.10/1M | $0.20/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| meta/muse-spark-1.3 | chat | 1.0M | $1.25/1M | $4.25/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| meta/muse-spark-1.3-contributor | chat | 1.0M | $0.10/1M | $0.20/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| meta/muse-spark-open | chat | 1.0M | $1.25/1M | $4.25/1M | fireworks, baseten, togetherai, modal, nebius | implicit-caching, reasoning, structured-output, tool-use, vision |
| meta/muse-wonder-1.0 | chat | 1.0M | $4.00/1M | $20.00/1M | meta | reasoning, tool-use, implicit-caching, file-input, vision, structured-output |
| microsoft/microsoft-decision-1 | evaluation | 32.8K | $0.04/1M | $0.00/1M | azure |  |
| microsoft/mai-image-2.6 | image | 0 | $5.00/1M | $38.00/1M | azure | image-generation |
| microsoft/mai-image-2.6-flash | image | 0 | $1.75/1M | $19.00/1M | azure | image-generation |
| microsoft/mai-transcribe-2 | transcription | 0 | $0.00/1M | $0.00/1M | azure |  |
| microsoft/mai-transcribe-2-streaming | transcription | 0 | $0.00/1M | $0.00/1M | azure | websockets |
| microsoft/mai-voice-2 | speech | 0 | $22.00/1M | $0.00/1M | azure |  |
| microsoft/mai-voice-2-flash | speech | 0 | $15.00/1M | $0.00/1M | azure |  |
| microsoft/mai-voice-2.1 | speech | 0 | $22.00/1M | $0.00/1M | azure |  |
| microsoft/mai-voice-2.1-flash | speech | 0 | $15.00/1M | $0.00/1M | azure |  |
| minimax/minimax-h3 | video | 0 | $0.00/1M | $0.00/1M | minimax | video-generation, image-to-video, text-to-video, reference-to-video, vision, open-weights |
| minimax/minimax-h3-max | video | 0 | $0.00/1M | $0.00/1M | minimax | video-generation, text-to-video, image-to-video, vision, open-weights |
| minimax/minimax-m2 | chat | 205K | $0.30/1M | $1.20/1M | minimax | reasoning, tool-use, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.1 | chat | 204.8K | $0.30/1M | $1.20/1M | minimax, bedrock | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.1-lightning | chat | 204.8K | $0.30/1M | $2.40/1M | minimax | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.5 | chat | 1M | $0.30/1M | $1.20/1M | minimax, bedrock | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.5-highspeed | chat | 204.8K | $0.60/1M | $2.40/1M | minimax | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.7 | chat | 204.8K | $0.21/1M | $0.84/1M | minimax, novita, gmicloud | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m2.7-highspeed | chat | 204.8K | $0.60/1M | $2.40/1M | minimax | reasoning, tool-use, implicit-caching, explicit-caching, structured-output, open-weights |
| minimax/minimax-m3 | chat | 1.0M | $0.24/1M | $0.96/1M | minimax, fireworks, nebius, gmicloud | implicit-caching, reasoning, tool-use, vision, file-input, structured-output, open-weights |
| minimax/minimax-m3.1-flash | chat | 1.0M | $0.10/1M | $0.50/1M | minimax, fireworks, togetherai, inference-net, parasail, novita, nebius, gmicloud | implicit-caching, reasoning, tool-use, vision, structured-output, open-weights |
| mistral/codestral-embed | embedding | 0 | $0.15/1M | $0.00/1M | mistral |  |
| mistral/ministral-14b | chat | 262.1K | $0.20/1M | $0.20/1M | mistral | file-input, tool-use, vision, structured-output, open-weights |
| mistral/ministral-3b | chat | 131.1K | $0.10/1M | $0.10/1M | mistral | tool-use, vision, open-weights |
| mistral/ministral-8b | chat | 262.1K | $0.15/1M | $0.15/1M | mistral | tool-use, vision, structured-output, open-weights |
| mistral/codestral | chat | 128K | $0.30/1M | $0.90/1M | mistral | tool-use, structured-output |
| mistral/mistral-embed | embedding | 0 | $0.10/1M | $0.00/1M | mistral |  |
| mistral/mistral-large-3 | chat | 262.1K | $0.50/1M | $1.50/1M | mistral | tool-use, vision, structured-output, open-weights |
| mistral/mistral-large-4 | chat | 524.3K | $1.36/1M | $4.18/1M | mistral | reasoning, structured-output, tool-use, vision |
| mistral/mistral-medium-3.5 | chat | 262.1K | $1.50/1M | $7.50/1M | mistral | reasoning, tool-use, vision, structured-output |
| mistral/mistral-nemo | chat | 131.1K | $0.03/1M | $0.03/1M | deepinfra | tool-use, structured-output, open-weights |
| mistral/mistral-small | chat | 262.1K | $0.15/1M | $0.60/1M | mistral | tool-use, vision, structured-output, open-weights |
| mistral/voxtral-mini-transcribe-3 | transcription | 65.5K | $0.00/1M | $0.00/1M | mistral |  |
| mistral/voxtral-small-transcribe-3 | transcription | 65.5K | $0.00/1M | $0.00/1M | mistral |  |
| mistral/voxtral-mini-tts-3 | speech | 4.1K | $16.00/1M | $0.00/1M | mistral |  |
| mixedbread/toast-1 | chat | 131K | $0.30/1M | $0.72/1M | mixedbread | implicit-caching, tool-use, structured-output |
| moonshotai/kimi-k2 | chat | 131.1K | $0.57/1M | $2.30/1M | novita | tool-use, open-weights |
| moonshotai/kimi-k2.5 | chat | 262.1K | $0.60/1M | $3.00/1M | novita, bedrock | reasoning, vision, tool-use, implicit-caching, open-weights |
| moonshotai/kimi-k2.6 | chat | 262.1K | $0.80/1M | $3.40/1M | moonshotai, novita | implicit-caching, reasoning, tool-use, vision, video-input, open-weights |
| moonshotai/kimi-k2.7-code | chat | 256K | $0.95/1M | $4.00/1M | moonshotai | reasoning, tool-use, implicit-caching, file-input, vision, video-input, structured-output, open-weights |
| moonshotai/kimi-k2.7-code-highspeed | chat | 262.1K | $1.90/1M | $8.00/1M | moonshotai | reasoning, tool-use, vision, file-input, implicit-caching, video-input, structured-output, open-weights |
| moonshotai/kimi-k3 | chat | 1.0M | $0.18/1M | $9.34/1M | moonshotai, fireworks, baseten, morph, nebius, togetherai, digitalocean, modal, wafer, alibaba, deepinfra, runware, parasail, boundless, inference-net, bedrock, relace | reasoning, tool-use, implicit-caching, file-input, vision, video-input, structured-output, open-weights |
| moonshotai/kimi-k3-fast | chat | 1M | $4.50/1M | $22.50/1M | fireworks, morph | reasoning, tool-use, implicit-caching, file-input, vision, structured-output, open-weights |
| moonshotai/kimi-k3.1-pro | chat | 1.0M | $2.00/1M | $10.00/1M | moonshotai | implicit-caching, reasoning, structured-output, tool-use, vision, open-weights |
| moonshotai/kimi-k3.1-pro-highspeed | chat | 1.0M | $4.00/1M | $20.00/1M | moonshotai | implicit-caching, reasoning, structured-output, tool-use, vision, open-weights |
| morph/morph-v3-fast | chat | 81.9K | $0.80/1M | $1.20/1M | morph | structured-output |
| morph/morph-v3-large | chat | 81.9K | $0.90/1M | $1.90/1M | morph | structured-output |
| nvidia/nemotron-3-nano-30b-a3b | chat | 262.1K | $0.05/1M | $0.20/1M | deepinfra | reasoning, tool-use, open-weights |
| nvidia/nemotron-3-ultra-550b-a55b | chat | 1M | $0.50/1M | $2.20/1M | deepinfra, baseten, fireworks | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| nvidia/nemotron-3.5-lightning | chat | 262.1K | $0.04/1M | $0.12/1M | runinfra, deepinfra, fireworks | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| nvidia/nemotron-3.5-lightning-free | chat | 1M | $0.00/1M | $0.00/1M | nvidia | reasoning, tool-use, structured-output, implicit-caching, open-weights, free |
| nvidia/nemotron-3-super-120b-a12b | chat | 256K | $0.15/1M | $0.65/1M | bedrock | reasoning, tool-use, structured-output, open-weights |
| nvidia/nemotron-3.5-super-vl | chat | 1.0M | $0.10/1M | $0.50/1M | baseten, togetherai, deepinfra, gmicloud | implicit-caching, reasoning, structured-output, tool-use, vision, open-weights |
| nvidia/nemotron-nano-12b-v2-vl | chat | 131.1K | $0.20/1M | $0.60/1M | bedrock | vision, reasoning, tool-use, open-weights |
| nvidia/nemotron-nano-9b-v2 | chat | 131.1K | $0.06/1M | $0.23/1M | bedrock | reasoning, tool-use, open-weights |
| openai/gpt-5.1-codex-max | chat | 400K | $1.25/1M | $10.00/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5.1-codex-mini | chat | 400K | $0.25/1M | $2.00/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5.1-thinking | chat | 400K | $1.25/1M | $10.00/1M | openai, azure | tool-use, implicit-caching, file-input, reasoning, vision, web-search, fast, structured-output |
| openai/gpt-5.2 | chat | 400K | $1.75/1M | $14.00/1M | azure, openai | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output, websockets |
| openai/gpt-5.2-pro | chat | 400K | $21.00/1M | $168.00/1M | openai | tool-use, vision, reasoning, file-input, web-search, structured-output |
| openai/gpt-5.2-codex | chat | 400K | $1.75/1M | $14.00/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5.3-codex | chat | 400K | $1.75/1M | $14.00/1M | openai, azure | reasoning, tool-use, file-input, vision, web-search, implicit-caching, fast, structured-output, websockets |
| openai/gpt-5.4 | chat | 1.1M | $2.50/1M | $15.00/1M | openai, azure | reasoning, tool-use, vision, file-input, implicit-caching, web-search, fast, structured-output, websockets |
| openai/gpt-5.4-mini | chat | 400K | $0.75/1M | $4.50/1M | openai, azure | reasoning, tool-use, vision, file-input, implicit-caching, web-search, fast, structured-output, websockets |
| openai/gpt-5.4-nano | chat | 400K | $0.20/1M | $1.25/1M | openai, azure | reasoning, tool-use, implicit-caching, web-search, vision, file-input, structured-output, websockets |
| openai/gpt-5.4-pro | chat | 1.1M | $30.00/1M | $180.00/1M | openai, azure | reasoning, tool-use, vision, file-input, web-search, structured-output |
| openai/gpt-5.5 | chat | 1M | $5.00/1M | $30.00/1M | openai, azure, bedrock | file-input, implicit-caching, reasoning, tool-use, vision, web-search, fast, structured-output, websockets |
| openai/gpt-5.5-pro | chat | 1M | $30.00/1M | $180.00/1M | openai | reasoning, tool-use, file-input, web-search, vision, structured-output, websockets |
| openai/gpt-5.6-luna | chat | 1.1M | $0.20/1M | $1.20/1M | openai, azure, bedrock | reasoning, file-input, tool-use, vision, implicit-caching, web-search, fast, explicit-caching, structured-output, websockets |
| openai/gpt-5.6-sol | chat | 1.1M | $4.00/1M | $20.00/1M | openai, azure, bedrock | reasoning, tool-use, implicit-caching, file-input, vision, web-search, fast, explicit-caching, structured-output, websockets |
| openai/gpt-5.6-terra | chat | 1.1M | $2.00/1M | $12.00/1M | openai, azure, bedrock | reasoning, web-search, file-input, tool-use, vision, implicit-caching, fast, explicit-caching, structured-output, websockets |
| openai/gpt-image-1-mini | image | 0 | $2.00/1M | $8.00/1M | openai | image-generation, implicit-caching |
| openai/gpt-image-1.5 | image | 0 | $5.00/1M | $32.00/1M | openai | image-generation, implicit-caching |
| openai/gpt-image-2 | image | 0 | $5.00/1M | $30.00/1M | openai | image-generation, implicit-caching |
| openai/gpt-image-2.5-flare | image | 0 | $5.00/1M | $30.00/1M | openai | image-generation, implicit-caching |
| openai/gpt-image-2.5-sunburst | image | 0 | $5.00/1M | $30.00/1M | openai | image-generation, implicit-caching |
| openai/gpt-live-transcribe | transcription | 0 | $0.00/1M | $0.00/1M | openai | websockets |
| openai/gpt-oss-120b | chat | 131.1K | $0.35/1M | $0.75/1M | baseten, bedrock, fireworks, groq, parasail, nebius, togetherai, cerebras | reasoning, tool-use, structured-output, open-weights |
| openai/gpt-oss-20b | chat | 131.1K | $0.07/1M | $0.30/1M | bedrock, groq, deepinfra, parasail | reasoning, tool-use, structured-output, open-weights |
| openai/gpt-oss-safeguard-120b | chat | 128K | $0.15/1M | $0.60/1M | bedrock | reasoning, tool-use, structured-output, open-weights |
| openai/gpt-oss-safeguard-20b | chat | 131.1K | $0.07/1M | $0.30/1M | groq, bedrock | implicit-caching, reasoning, tool-use, structured-output, open-weights |
| openai/gpt-transcribe | transcription | 0 | $0.00/1M | $0.00/1M | openai | websockets |
| openai/gpt-4.1 | chat | 1.0M | $2.00/1M | $8.00/1M | azure, openai | file-input, implicit-caching, tool-use, vision, web-search, structured-output |
| openai/gpt-4.1-mini | chat | 1.0M | $0.40/1M | $1.60/1M | azure, openai | file-input, implicit-caching, tool-use, vision, web-search, structured-output |
| openai/gpt-4.1-nano | chat | 1.0M | $0.10/1M | $0.40/1M | azure | file-input, implicit-caching, tool-use, vision, web-search, structured-output |
| openai/gpt-4o | chat | 128K | $2.50/1M | $10.00/1M | azure, openai | file-input, implicit-caching, tool-use, vision, web-search, structured-output |
| openai/gpt-4o-mini | chat | 128K | $0.15/1M | $0.60/1M | azure, openai | file-input, implicit-caching, tool-use, vision, web-search, structured-output |
| openai/gpt-4o-mini-transcribe | transcription | 0 | $1.25/1M | $5.00/1M | openai |  |
| openai/gpt-4o-transcribe | transcription | 0 | $2.50/1M | $10.00/1M | openai |  |
| openai/gpt-5 | chat | 400K | $1.25/1M | $10.00/1M | azure, openai | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5-mini | chat | 400K | $0.25/1M | $2.00/1M | azure, openai | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5-nano | chat | 400K | $0.05/1M | $0.40/1M | azure, openai | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5-pro | chat | 400K | $15.00/1M | $120.00/1M | openai | file-input, reasoning, tool-use, vision, web-search |
| openai/gpt-5-codex | chat | 400K | $1.25/1M | $10.00/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-5.1-codex | chat | 400K | $1.25/1M | $10.00/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/gpt-6-astra | chat | 1.1M | $10.00/1M | $50.00/1M | openai, azure, bedrock | fast, file-input, implicit-caching, reasoning, tool-use, vision, web-search, explicit-caching, structured-output, websockets |
| openai/gpt-6-luna | chat | 1.1M | $0.10/1M | $0.50/1M | openai, azure, bedrock | explicit-caching, fast, file-input, implicit-caching, reasoning, structured-output, tool-use, vision, web-search |
| openai/gpt-6-luna-decisions | evaluation | 1.1M | $0.10/1M | $0.00/1M | openai |  |
| openai/gpt-6-sol | chat | 1.1M | $2.00/1M | $10.00/1M | openai, azure, bedrock | explicit-caching, fast, file-input, implicit-caching, reasoning, structured-output, tool-use, vision, web-search |
| openai/gpt-6.1-sol | chat | 1.1M | $2.00/1M | $10.00/1M | openai, bedrock, azure | explicit-caching, fast, file-input, implicit-caching, reasoning, structured-output, tool-use, vision, web-search |
| openai/gpt-6.1-terra | chat | 1.1M | $1.00/1M | $5.00/1M | openai, azure | explicit-caching, fast, file-input, implicit-caching, reasoning, structured-output, tool-use, vision, web-search |
| openai/gpt-live-1 | realtime | 0 | $0.00/1M | $0.00/1M | openai | audio-input |
| openai/gpt-realtime-mini | realtime | 0 | $0.60/1M | $2.40/1M | openai |  |
| openai/gpt-realtime-1.5 | realtime | 0 | $4.00/1M | $16.00/1M | openai |  |
| openai/gpt-realtime-2 | realtime | 0 | $4.00/1M | $24.00/1M | openai | websockets |
| openai/gpt-realtime-2.1 | realtime | 128K | $4.00/1M | $24.00/1M | openai |  |
| openai/gpt-realtime-whisper | transcription | 0 | $0.00/1M | $0.00/1M | openai | websockets |
| openai/o1 | chat | 200K | $15.00/1M | $60.00/1M | azure | file-input, reasoning, tool-use, vision, implicit-caching, structured-output |
| openai/o3 | chat | 200K | $2.00/1M | $8.00/1M | openai | file-input, implicit-caching, reasoning, tool-use, vision, web-search, fast, structured-output |
| openai/o3-pro | chat | 200K | $20.00/1M | $80.00/1M | openai | reasoning, vision, file-input, tool-use, web-search, structured-output |
| openai/o3-mini | chat | 200K | $1.10/1M | $4.40/1M | azure | implicit-caching, reasoning, tool-use, structured-output |
| openai/o4-mini | chat | 200K | $1.10/1M | $4.40/1M | azure | file-input, implicit-caching, reasoning, tool-use, vision, web-search, structured-output |
| openai/text-embedding-3-large | embedding | 0 | $0.13/1M | $0.00/1M | azure, openai |  |
| openai/text-embedding-3-small | embedding | 0 | $0.02/1M | $0.00/1M | azure, openai |  |
| openai/text-embedding-ada-002 | embedding | 0 | $0.10/1M | $0.00/1M | azure, openai |  |
| openai/tts-1 | speech | 0 | $15.00/1M | $0.00/1M | openai | speech |
| openai/tts-1-hd | speech | 0 | $30.00/1M | $0.00/1M | openai | speech |
| openai/whisper-1 | transcription | 0 | $0.00/1M | $0.00/1M | openai | translation, transcription, open-weights |
| particle/particle-1 | chat | 1.0M | $0.13/1M | $0.50/1M | particle | implicit-caching, reasoning, tool-use, vision |
| perplexity/pplx-decider-flash-v1.2 | evaluation | 262.1K | $0.01/1M | $0.00/1M | perplexity | vision |
| perplexity/pplx-decider-v1.2 | evaluation | 262.1K | $0.02/1M | $0.00/1M | perplexity | vision |
| perplexity/pplx-embed-v1-0.6b | embedding | 32K | $0.00/1M | $0.00/1M | perplexity | open-weights |
| perplexity/pplx-embed-v1-4b | embedding | 32K | $0.03/1M | $0.00/1M | perplexity | open-weights |
| perplexity/sonar | chat | 127K | $0.25/1M | $2.50/1M | perplexity | vision, web-search, structured-output, reasoning |
| poolside/laguna-s-2.1 | chat | 1M | $0.09/1M | $0.18/1M | poolside | reasoning, tool-use, open-weights |
| poolside/laguna-s-2.1-free | chat | 256K | $0.00/1M | $0.00/1M | poolside | reasoning, tool-use, open-weights, free |
| prodia/flux-fast-schnell | image | 512 | $0.00/1M | $0.00/1M | prodia | image-generation, open-weights |
| quiverai/arrow-2 | chat | 131.1K | $4.00/1M | $20.00/1M | quiverai | tool-use, reasoning, vision, image-generation, implicit-caching |
| quiverai/arrow-2-telos | chat | 131.1K | $6.00/1M | $30.00/1M | quiverai | tool-use, reasoning, vision, image-generation, implicit-caching |
| recraft/recraft-v2 | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v3 | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4 | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4-pro | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4.1 | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4.1-flash | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4.1-pro | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4.1-utility | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| recraft/recraft-v4.1-utility-pro | image | 0 | $0.00/1M | $0.00/1M | recraft | image-generation |
| reflection/beam | chat | 1.0M | $0.25/1M | $1.00/1M | reflection, baseten, togetherai, nebius | implicit-caching, reasoning, structured-output, tool-use, open-weights |
| sakana/fugu-max | chat | 1M | $2.00/1M | $6.00/1M | sakana | reasoning, tool-use, vision, implicit-caching, structured-output |
| sakana/fugu-ultra | chat | 1M | $5.00/1M | $30.00/1M | sakana | vision, tool-use, reasoning, structured-output |
| sakana/fugu-ultra-v2 | chat | 1M | $5.00/1M | $30.00/1M | sakana | reasoning, tool-use, vision, implicit-caching, structured-output |
| sakana/namazu | chat | 256K | $0.95/1M | $4.00/1M | sakana | vision, file-input, implicit-caching, reasoning, structured-output, tool-use, web-search |
| scaledown/scaledown-classify-1 | evaluation | 0 | $0.05/1M | $0.00/1M | scaledown |  |
| scaledown/scaledown-extract-1 | chat | 1M | $0.05/1M | $0.00/1M | scaledown | structured-output |
| scaledown/scaledown-summarize-1 | chat | 1M | $0.05/1M | $0.00/1M | scaledown |  |
| stealth/glyph-cluster | chat | 256K | $0.00/1M | $0.00/1M | stealth | reasoning, tool-use, implicit-caching, free |
| stepfun/step-3.7-flash | chat | 256K | $0.20/1M | $1.15/1M | stepfun | implicit-caching, reasoning, tool-use, vision, structured-output, open-weights |
| stepfun/step-5-preview | chat | 1.0M | $1.00/1M | $2.70/1M | stepfun, deepinfra, tencent | implicit-caching, vision, structured-output, open-weights |
| tencent/hy3 | chat | 262.1K | $0.13/1M | $0.53/1M | deepinfra, novita, gmicloud, tencent | implicit-caching, reasoning, tool-use, structured-output, open-weights |
| tencent/hy-mt2-lite | chat | 8K | $0.04/1M | $0.18/1M | tencent | open-weights |
| tencent/hy-mt2-plus | chat | 8K | $0.07/1M | $0.29/1M | tencent | structured-output, open-weights |
| tencent/hy-mt2-pro | chat | 8K | $0.07/1M | $0.29/1M | tencent | structured-output, open-weights |
| tencent/hy4-preview | chat | 1.0M | $0.83/1M | $2.50/1M | tencent | reasoning, tool-use, implicit-caching, open-weights |
| thinkingmachines/inkling | chat | 524.3K | $0.95/1M | $4.05/1M | togetherai, modal, thinkingmachines, deepinfra | reasoning, tool-use, vision, file-input, structured-output, open-weights |
| thinkingmachines/inkling-small | chat | 1M | $0.30/1M | $1.20/1M | deepinfra, thinkingmachines | reasoning, tool-use, vision, file-input, implicit-caching, structured-output, open-weights |
| topaz/proteus | video | 0 | $0.00/1M | $0.00/1M | topaz | video-editing |
| topaz/starlight-precise-2.6 | video | 0 | $0.00/1M | $0.00/1M | topaz | video-editing |
| topaz/wonder-3.5 | image | 0 | $0.00/1M | $0.00/1M | topaz | image-editing |
| typesafe-ai/jev | evaluation | 32K | $0.04/1M | $0.00/1M | typesafe-ai, digitalocean |  |
| typesafe-ai/jev-1.13.0 | evaluation | 32K | $0.04/1M | $0.00/1M | typesafe-ai | vision |
| typesafe-ai/jev-1.14.0 | evaluation | 32K | $0.04/1M | $0.00/1M | typesafe-ai | vision |
| voyage/voyage-3.5 | embedding | 0 | $0.06/1M | $0.00/1M | voyage |  |
| voyage/voyage-3.5-lite | embedding | 0 | $0.02/1M | $0.00/1M | voyage |  |
| voyage/voyage-4 | embedding | 32K | $0.06/1M | $0.00/1M | voyage |  |
| voyage/voyage-4-large | embedding | 32K | $0.12/1M | $0.00/1M | voyage |  |
| voyage/voyage-4-lite | embedding | 32K | $0.02/1M | $0.00/1M | voyage |  |
| voyage/voyage-code-2 | embedding | 0 | $0.12/1M | $0.00/1M | voyage |  |
| voyage/voyage-code-3 | embedding | 0 | $0.18/1M | $0.00/1M | voyage |  |
| voyage/voyage-finance-2 | embedding | 0 | $0.12/1M | $0.00/1M | voyage |  |
| voyage/voyage-law-2 | embedding | 0 | $0.12/1M | $0.00/1M | voyage |  |
| voyage/rerank-2.5 | reranking | 32K | $0.05/1M | $0.00/1M | voyage | reranking |
| voyage/rerank-2.5-lite | reranking | 32K | $0.02/1M | $0.00/1M | voyage | reranking |
| voyage/rerank-3 | reranking | 32K | $0.05/1M | $0.00/1M | voyage | reranking |
| voyage/rerank-3-lite | reranking | 32K | $0.02/1M | $0.00/1M | voyage | reranking |
| voyage/voyage-3-large | embedding | 0 | $0.18/1M | $0.00/1M | voyage |  |
| spacexai/grok-4.20-non-reasoning-beta | chat | 2M | $1.25/1M | $2.50/1M | xai | tool-use, implicit-caching, vision, file-input, web-search, structured-output |
| spacexai/grok-4.20-reasoning-beta | chat | 2M | $1.25/1M | $2.50/1M | xai | reasoning, tool-use, vision, file-input, implicit-caching, web-search, structured-output |
| spacexai/grok-4.20-multi-agent-beta | chat | 2M | $1.25/1M | $2.50/1M | xai | reasoning, tool-use, implicit-caching, vision, file-input, web-search, structured-output |
| spacexai/grok-4.20-multi-agent | chat | 2M | $1.25/1M | $2.50/1M | xai | reasoning, tool-use, implicit-caching, vision, file-input, web-search, structured-output |
| spacexai/grok-4.20-non-reasoning | chat | 2M | $1.25/1M | $2.50/1M | xai, vertex | tool-use, implicit-caching, vision, file-input, web-search, structured-output |
| spacexai/grok-4.20-reasoning | chat | 2M | $1.25/1M | $2.50/1M | xai, vertex | reasoning, vision, tool-use, file-input, implicit-caching, web-search, structured-output |
| spacexai/grok-4.3 | chat | 1M | $1.25/1M | $2.50/1M | xai, vertex | reasoning, tool-use, implicit-caching, file-input, vision, web-search, structured-output |
| spacexai/grok-4.5 | chat | 500K | $2.00/1M | $6.00/1M | xai | reasoning, tool-use, implicit-caching, file-input, vision, web-search, structured-output |
| spacexai/grok-4.6 | chat | 500K | $2.00/1M | $6.00/1M | xai | reasoning, tool-use, implicit-caching, vision, structured-output |
| spacexai/grok-4.7 | chat | 500K | $2.00/1M | $6.00/1M | xai | reasoning, tool-use, implicit-caching, vision, structured-output |
| spacexai/grok-build-0.1 | chat | 256K | $1.00/1M | $2.00/1M | xai | reasoning, implicit-caching, vision, tool-use, web-search, structured-output |
| spacexai/grok-imagine-video | video | 0 | $0.00/1M | $0.00/1M | xai | image-to-video, text-to-video, video-editing, audio-generation |
| spacexai/grok-imagine-image | image | 0 | $0.00/1M | $0.00/1M | xai | image-generation |
| spacexai/grok-imagine-image-2.0 | image | 0 | $0.00/1M | $0.00/1M | xai | image-generation |
| spacexai/grok-imagine-video-1.5 | video | 0 | $0.00/1M | $0.00/1M | xai | image-to-video, reference-to-video, text-to-video |
| spacexai/grok-imagine-video-1.5-lite | video | 0 | $0.00/1M | $0.00/1M | xai | image-to-video, text-to-video |
| spacexai/grok-stt | transcription | 0 | $0.00/1M | $0.00/1M | xai | websockets |
| spacexai/grok-tts | speech | 0 | $15.00/1M | $0.00/1M | xai |  |
| spacexai/grok-voice-3.0 | realtime | 0 | $0.00/1M | $0.00/1M | xai | audio-input, reasoning, tool-use, websockets |
| spacexai/grok-voice-think-fast-2.0 | realtime | 0 | $0.00/1M | $0.00/1M | xai |  |
| xiaomi/mimo-v2.5 | chat | 1.1M | $0.12/1M | $0.24/1M | gmicloud | reasoning, tool-use, implicit-caching, vision, structured-output, open-weights |
| xiaomi/mimo-v2.5-pro | chat | 1.1M | $0.30/1M | $0.61/1M | gmicloud | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| xiaomi/mimo-v2.6-flash | chat | 1.1M | $0.12/1M | $0.24/1M | xiaomi, deepinfra, novita, gmicloud, boundless, inference-net, tencent | audio-input, implicit-caching, reasoning, structured-output, tool-use, video-input, vision, open-weights |
| xiaomi/mimo-v2.6-pro | chat | 1.1M | $0.37/1M | $0.74/1M | xiaomi, deepinfra, novita, gmicloud, boundless, tencent | audio-input, implicit-caching, reasoning, structured-output, tool-use, video-input, vision, open-weights |
| xiaomi/mimo-v2.6-pro-ultraspeed | chat | 1.0M | $4.35/1M | $8.70/1M | xiaomi | audio-input, implicit-caching, reasoning, structured-output, tool-use, video-input, vision, open-weights |
| zai/glm-4.5 | chat | 128K | $0.60/1M | $2.20/1M | zai | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-4.5-air | chat | 128K | $0.20/1M | $1.10/1M | zai | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-4.5v | chat | 66K | $0.60/1M | $1.80/1M | novita, zai | implicit-caching, reasoning, tool-use, vision, structured-output, open-weights |
| zai/glm-4.6 | chat | 204.8K | $0.60/1M | $2.20/1M | zai, deepinfra, novita | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-4.7 | chat | 204.8K | $0.40/1M | $1.75/1M | zai, novita, deepinfra, bedrock | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-4.7-flash | chat | 200K | $0.07/1M | $0.40/1M | zai, bedrock | implicit-caching, reasoning, tool-use, structured-output, open-weights |
| zai/glm-4.7-flashx | chat | 200K | $0.06/1M | $0.40/1M | zai | reasoning, tool-use, implicit-caching, open-weights |
| zai/glm-5 | chat | 202.8K | $1.00/1M | $3.20/1M | zai, bedrock, novita | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5-turbo | chat | 202.8K | $1.20/1M | $4.00/1M | zai | reasoning, tool-use, implicit-caching |
| zai/glm-5.1 | chat | 204.8K | $1.38/1M | $4.40/1M | zai, novita | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5.2 | chat | 1.0M | $0.19/1M | $1.80/1M | zai, baseten, wafer, deepinfra, parasail, togetherai, streamlake, novita, alibaba, digitalocean, nebius, friendli, runware, gmicloud, inceptron, tencent | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5.2-fast | chat | 1M | $2.10/1M | $6.60/1M | baseten, alibaba | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5.3 | chat | 1.0M | $0.07/1M | $1.40/1M | zai, deepinfra, friendli, baseten, gmicloud, fireworks, modal, parasail, togetherai, digitalocean, morph, novita, wafer, runware, streamlake, inceptron, crusoe, alibaba, inference-net, relace, boundless, bedrock, tencent, runinfra | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5.3-fast | chat | 1.0M | $2.10/1M | $6.60/1M | baseten, alibaba | reasoning, tool-use, implicit-caching, structured-output, open-weights |
| zai/glm-5.3-flash | chat | 1.0M | $0.04/1M | $0.07/1M | zai, baseten, novita, gmicloud, deepinfra, runinfra, wafer, modal, togetherai, morph, parasail, friendli, digitalocean, fireworks, relace, runware, particle, streamlake, boundless, inference-net, tencent, pareto | reasoning, tool-use, implicit-caching, vision, structured-output, open-weights |
| zai/glm-5.3-flashx | chat | 1M | $0.37/1M | $1.25/1M | zai | reasoning, tool-use, implicit-caching, vision, structured-output, open-weights |
| zai/glm-5v-turbo | chat | 200K | $1.20/1M | $4.00/1M | zai | reasoning, tool-use, implicit-caching, vision, file-input, structured-output |
| exa/search | tool | 0 | $0.00/1M | $0.00/1M | exa | web-search, tool |
| perplexity/search | tool | 0 | $0.00/1M | $0.00/1M | perplexity | web-search, tool |
| parallel/search | tool | 0 | $0.00/1M | $0.00/1M | parallel | web-search, tool |
| tako/search | tool | 0 | $0.00/1M | $0.00/1M | tako | web-search, tool |
| browserbase/search | tool | 0 | $0.00/1M | $0.00/1M | browserbase | web-search, tool |
| browserbase/fetch | tool | 0 | $0.00/1M | $0.00/1M | browserbase | tool |

---

## Detailed Model Information

### alibaba/qwen-3-32b

Qwen3-32B is a world-class model with comparable quality to DeepSeek R1 while outperforming GPT-4.1 and Claude Sonnet 3.7. It excels in code-gen, tool-calling, and advanced reasoning, making it an exceptional model for a wide range of production use cases.

- **Max output tokens:** 16.4K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen-3-32b

### alibaba/qwen3-coder-30b-a3b

Efficient coding specialist balancing performance with cost-effectiveness for daily development tasks while maintaining strong tool integration capabilities.

- **Max output tokens:** 8.2K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-coder-30b-a3b

### alibaba/qwen3-max-thinking

Compared with the snapshot as of September 23, 2025, the Qwen-3 series Max model in this release achieves an effective integration of thinking and non-thinking modes, resulting in a comprehensive and substantial improvement in the model’s overall performance. In thinking mode, the model simultaneously supports web search, web information extraction, and a code interpreter tool, enabling it to tackle more complex and challenging problems with greater accuracy by leveraging external tools while engaging in slow, deliberative reasoning. This version is based on a snapshot taken on January 23, 2026.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.24/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-max-thinking

### alibaba/qwen3-vl-235b-a22b-instruct



- **Max output tokens:** 262.1K
- **Cached input cost:** $0.11/1M
- **Details:** https://vercel.com/ai-gateway/models/qwen3-vl-235b-a22b-instruct

### alibaba/qwen3.5-flash

The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the 3 series, these models deliver a leap forward in performance for both pure text and multimodal tasks, offering fast response times while balancing inference speed and overall performance.

- **Max output tokens:** 64K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.5-flash

### alibaba/qwen3.5-plus

The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In a variety of task evaluations, the 3.5 series consistently demonstrates performance on par with state-of-the-art leading models. Compared to the 3 series, these models show a leap forward in both pure-text and multimodal capabilities.

- **Max output tokens:** 64K
- **Cached input cost:** $0.04/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.5-plus

### alibaba/qwen3.6-27b

The Qwen3.6 35B-A3B native vision-language model is built on a hybrid architecture that integrates linear attention mechanisms with a sparse mixture-of-experts framework, achieving higher inference efficiency. Compared with the 3.5-35B-A3B, this model demonstrates significantly improved agentic coding capabilities, mathematical and code reasoning abilities, spatial intelligence, as well as object localization and object detection performance.

- **Max output tokens:** 256K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.6-27b

### alibaba/qwen3.6-plus

The Qwen3.6 native vision-language Plus series models demonstrate exceptional performance on par with the current state-of-the-art models, with a significant improvement in overall results compared to the 3.5 series. The models have been markedly enhanced in code-related capabilities such as agentic coding, front-end programming, and Vibe coding, as well as in multi-modal general object recognition, OCR, and object localization.

- **Max output tokens:** 64K
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.6-plus

### alibaba/qwen3.7-flash

The Qwen3.7 native vision-language Flash model series delivers a comprehensive upgrade over 3.6-Flash in multimodal understanding and agent execution. This model particularly excels in enhanced multimodal foundations with stronger universal object recognition, further improved real-world perception and spatial intelligence, significantly upgraded multimodal agent capabilities for Search Agent and CI Agent scenarios with more stable end-to-end task execution, as well as optimized multimodal coding for a smoother vibe coding experience.

- **Max output tokens:** 64K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.7-flash

### alibaba/qwen3.7-max

Qwen3.7 is a next‑generation flagship model designed for the agent‑centric era, with its core strengths lying in the breadth and depth of its agent‑level capabilities: it excels at programming, office and productivity tasks, and long‑term autonomous execution.

- **Max output tokens:** 64K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.7-max

### alibaba/qwen3.7-plus

Among the Qwen3.7 series, the cost-effective Plus model builds on its robust text capabilities while delivering a comprehensive upgrade to its vision‑language abilities, all while preserving its full‑stack agent‑level intelligence for coding, tool use, and productivity workflows.

- **Max output tokens:** 64K
- **Cached input cost:** $0.08/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.7-plus

### alibaba/qwen3.8-27b

Alibaba's 27B dense multimodal model for agentic coding, tool use, research, and long-running workflows.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-27b

### alibaba/qwen3.8-flash

Qwen3.8-Flash is Qwen’s fast, cost-efficient multimodal model, combining advanced reasoning and generation with a native 1M-token context window. Built for coding, agentic workflows, and visual understanding, it handles large codebases, long documents, charts, videos, and desktop applications.

- **Max output tokens:** 128K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-flash

### alibaba/qwen3.8-max

Qwen 3.8 Max is a 2.4-trillion-parameter MoE model delivering a comprehensive leap in coding and professional work. Autonomously codes and delivers complete projects spanning 10+ days. Handles hundreds of specialized tasks across legal, financial, design, and other professional domains, producing production-grade results end-to-end in a single conversation. Native visual understanding runs through the full cycle of planning, execution, and verification, enabling deep semantic analysis of ultra-long documents and extended video content. In long-horizon tasks, plans autonomously, iterates through closed feedback loops, and continuously evolves.

- **Max output tokens:** 128K
- **Cached input cost:** $0.25/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-max

### alibaba/qwen3.8-max-prime

Qwen 3.8 Max Prime is the high-speed edition of Qwen 3.8 Max, retaining its 2.4-trillion-parameter MoE architecture and full capabilities while delivering 1.5–2× higher output throughput. Built for coding, office automation, and long-running agent workflows, it supports autonomous multi-day development, professional knowledge work, and visual understanding with a 1M-token context window.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-max-prime

### alibaba/qwen3.8-omni-flash

Qwen3.8-Omni-Flash is Alibaba’s native multimodal model for understanding text, images, audio, and video and generating text. Built on Qwen3.8-Flash-Next, it supports a 1M-token context window, reasoning, and tool calling for coding, knowledge work, and multimedia analysis.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-omni-flash

### alibaba/qwen-3-235b

Qwen3-235B-A22B-Instruct-2507 is the updated version of the Qwen3-235B-A22B non-thinking mode, featuring Significant improvements in general capabilities, including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage.

- **Max output tokens:** 16.4K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen-3-235b

### alibaba/qwen3-coder

Mixture-of-experts LLM with advanced coding and reasoning capabilities

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-coder

### alibaba/qwen3-coder-next

Qwen3-Coder-Next is an open-weight language model built specifically for coding, with strong performance on large-scale software engineering and agentic coding benchmarks. It uses a hybrid Mixture-of-Experts architecture to offer high capability at relatively modest active parameter counts, improving efficiency for real-world deployments. The model is trained on diverse code and natural language data so it can handle tasks like code generation, refactoring, debugging, repository-level reasoning, and technical explanation across multiple programming languages. It is also optimized for tool use and function calling, making it suitable as the core of coding agents that interact with shells, editors, issue trackers, and other developer tools.

- **Max output tokens:** 256K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-coder-next

### alibaba/qwen3-embedding-0.6b

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).

- **Max output tokens:** 32.8K
- **Details:** https://vercel.com/ai-gateway/models/qwen3-embedding-0.6b

### alibaba/qwen3-embedding-4b

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).

- **Max output tokens:** 32.8K
- **Details:** https://vercel.com/ai-gateway/models/qwen3-embedding-4b

### alibaba/qwen3-embedding-8b

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).

- **Max output tokens:** 41K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-embedding-8b

### alibaba/qwen3-next-80b-a3b-instruct

A new generation of open-source, non-thinking mode model powered by Qwen3. This version demonstrates superior Chinese text understanding, augmented logical reasoning, and enhanced capabilities in text generation tasks over the previous iteration (Qwen3-235B-A22B-Instruct-2507).

- **Max output tokens:** 262.1K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-next-80b-a3b-instruct

### alibaba/qwen3-next-80b-a3b-thinking

Qwen3-Next uses a highly sparse MoE design: 80B total parameters, but only ~3B activated per inference step. Experiments show that, with global load balancing, increasing total expert parameters while keeping activated experts fixed steadily reduces training loss.Compared to Qwen3’s MoE (128 total experts, 8 routed), Qwen3-Next expands to 512 total experts, combining 10 routed experts + 1 shared expert — maximizing resource usage without hurting performance. The Qwen3-Next-80B-A3B-Thinking excels at complex reasoning tasks — outperforming higher-cost models like Qwen3-30B-A3B-Thinking-2507 and Qwen3-32B-Thinking, outpeforming the closed-source Gemini-2.5-Flash-Thinking on multiple benchmarks, and approaching the performance of our top-tier model Qwen3-235B-A22B-Thinking-2507.

- **Max output tokens:** 262.1K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3-next-80b-a3b-thinking

### alibaba/qwen3-vl-instruct

The Qwen3 series VL models has been comprehensively upgraded in areas such as visual coding and spatial perception. Its visual perception and recognition capabilities have significantly improved, supporting the understanding of ultra-long videos, and its OCR functionality has undergone a major enhancement.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.11/1M
- **Details:** https://vercel.com/ai-gateway/models/qwen3-vl-instruct

### alibaba/qwen-3-14b

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support

- **Max output tokens:** 16.4K
- **Details:** https://vercel.com/ai-gateway/models/qwen-3-14b

### alibaba/qwen-3-30b

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support

- **Max output tokens:** 16.4K
- **Details:** https://vercel.com/ai-gateway/models/qwen-3-30b

### alibaba/qwen3.8-2.4t-a95b

Open-weights release of the Qwen3.8 flagship (2.4T MoE, ~95B active); thinking always on with reasoning_effort low/medium/xhigh. The hosted Qwen 3.8 Max (vision, non-thinking, 1M default context) is Alibaba-only.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-2.4t-a95b

### alibaba/qwen3.8-max-0902

Qwen3.8-Max-0902 is an upgraded snapshot of qwen3.8-max.

- **Max output tokens:** 128K
- **Cached input cost:** $0.25/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/qwen3.8-max-0902

### alibaba/wan-v2.5-t2v-preview



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.5-t2v-preview

### alibaba/wan-v2.6-i2v



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.6-i2v

### alibaba/wan-v2.6-i2v-flash



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.6-i2v-flash

### alibaba/wan-v2.6-r2v



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.6-r2v

### alibaba/wan-v2.6-r2v-flash



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.6-r2v-flash

### alibaba/wan-v2.6-t2v



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.6-t2v

### alibaba/wan-v2.7-r2v



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.7-r2v

### alibaba/wan-v2.7-t2v



- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v2.7-t2v

### alibaba/wan-v3.0-video

All-in-one video generation model supporting text-to-video, image-to-video, first/last-frame, and omni-modal reference-based generation (image, video, and audio references) with synchronized audio, at up to 30 seconds per clip.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v3.0-video

### alibaba/wan-v3.0-video-prime

Wan 3.0 Video Prime is Alibaba’s high-speed, all-in-one AI video generation model for creating polished clips from text, images, video, and audio references. It produces videos up to 30 seconds long with synchronized dialogue, music, and sound effects, while accelerated generation makes it ideal for rapid creative iteration and production workflows.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/wan-v3.0-video-prime

### amazon/nova-2-lite

Nova 2 Lite is a fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text.

- **Max output tokens:** 1M
- **Cached input cost:** $0.08/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nova-2-lite

### amazon/nova-lite

A very low cost multimodal model that is lightning fast for processing image, video, and text inputs.

- **Max output tokens:** 8.2K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nova-lite

### amazon/nova-micro

A text-only model that delivers the lowest latency responses at very low cost.

- **Max output tokens:** 8.2K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nova-micro

### amazon/nova-pro

A highly capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks.

- **Max output tokens:** 8.2K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nova-pro

### amazon/titan-embed-text-v2

Amazon Titan Text Embeddings V2 is a light weight, efficient multilingual embedding model supporting 1024, 512, and 256 dimensions.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/titan-embed-text-v2

### anthropic/claude-fable-5

Claude Fable 5 is a Mythos-class model with robust safeguards. It can handle long-running, complex, and asynchronous tasks where previous models would have needed more frequent check-ins.

- **Max output tokens:** 128K
- **Cached input cost:** $1.00/1M
- **Details:** https://vercel.com/ai-gateway/models/claude-fable-5

### anthropic/claude-fable-5.1

Claude Fable 5.1 improves on Fable 5 across long-running agentic coding, knowledge work, and research, following instructions precisely over sessions that run unattended for hours.

- **Max output tokens:** 128K
- **Cached input cost:** $0.25/1M
- **Details:** https://vercel.com/ai-gateway/models/claude-fable-5.1

### anthropic/claude-haiku-4.5

Claude Haiku 4.5 matches Sonnet 4's performance on coding, computer use, and agent tasks at substantially lower cost and faster speeds. It delivers near-frontier performance and Claude’s unique character at a price point that works for scaled sub-agent deployments, free tier products, and intelligence-sensitive applications with budget constraints.

- **Max output tokens:** 64K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-haiku-4.5

### anthropic/claude-haiku-5.5

Claude Haiku 5.5 brings adaptive reasoning to Anthropic’s low-cost Haiku family for everyday tasks, tool use, and scaled sub-agent deployments. It is the first Haiku with adjustable effort levels, letting developers balance reasoning depth and token usage for each task, with the option to turn thinking off for simpler requests.

- **Max output tokens:** 128K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-haiku-5.5

### anthropic/claude-opus-4.5

Claude Opus 4.5 is Anthropic’s latest model in the Opus series, meant for demanding reasoning tasks and complex problem solving. This model has improvements in general intelligence and vision compared to previous iterations. In addition, it is suited for difficult coding tasks and agentic workflows, especially those with computer use and tool use, and can effectively handle context usage and external memory files.

- **Max output tokens:** 64K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-4.5

### anthropic/claude-opus-4.6

Opus 4.6 is the world’s best model for coding and professional work, built to power agents that take on whole categories of real-world work. It excels across the entire SDLC, breaking through on hard problems, identifying complex bugs, and demonstrating deeper codebase understanding. It also delivers a step-change in knowledge work, with near-production-ready documents, presentations, and spreadsheets on the first pass.

- **Max output tokens:** 128K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-4.6

### anthropic/claude-opus-4.7

Opus 4.7 builds on the coding and agentic strengths of Opus 4.6 with stronger performance on complex, multi-step tasks and more reliable agentic execution. It also brings improved performance on knowledge work, from drafting documents to building presentations and analyzing data.

- **Max output tokens:** 128K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-4.7

### anthropic/claude-opus-4.8

Opus 4.8 is a focused upgrade to Opus 4.7 and is Anthropic's best generally available model for coding, agentic tasks, and enterprise workflows. It builds on the strengths of previous Opus models with stronger performance on complex, multi-step coding tasks. Anthropic recommends using it on long-horizon coding and agentic tasks. It is also stronger on professional work, including document drafting, data analysis, and presentations.

- **Max output tokens:** 128K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-4.8

### anthropic/claude-opus-5

Claude Opus 5 is the latest model in Anthropic's Opus family and a step-change improvement over Opus 4.8. It delivers major gains over Opus 4.8 in agentic coding, professional knowledge work, and long-horizon reasoning, and it is stronger per token across effort levels.

- **Max output tokens:** 128K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-5

### anthropic/claude-opus-5.5

Claude Opus 5.5 is a step-change improvement over Opus 5 for agentic coding, long-running agentic tasks, knowledge work, communication, and vision. It is more token-efficient per completed task and reports progress, findings, and next steps in clear language.

- **Max output tokens:** 128K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-opus-5.5

### anthropic/claude-sonnet-4.5

Claude Sonnet 4.5 is the newest model in the Sonnet series, offering improvements and updates over Sonnet 4.

- **Max output tokens:** 64K
- **Cached input cost:** $0.30/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-sonnet-4.5

### anthropic/claude-sonnet-4.6

Claude Sonnet 4.6 is the most capable Sonnet-class model yet, with frontier performance across coding, agents, and professional work. It excels at iterative development, complex codebase navigation, end-to-end project management with memory, polished document creation, and confident computer use for web QA and workflow automation.

- **Max output tokens:** 128K
- **Cached input cost:** $0.30/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-sonnet-4.6

### anthropic/claude-sonnet-5

Sonnet 5 is an upgrade to Sonnet 4.6, with gains across agentic coding and professional work. It builds on the strengths of previous Sonnet models, bringing top-tier intelligence at Sonnet pricing for coding, agents, and everyday professional work at scale.

- **Max output tokens:** 128K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-sonnet-5

### anthropic/claude-sonnet-5.5

Claude Sonnet 5.5 is a step-change improvement over Sonnet 5 for well-scoped everyday work like building features, fixing bugs, and creating polished documents, slides, and spreadsheets. It also writes more clearly than Sonnet 5.

- **Max output tokens:** 128K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/claude-sonnet-5.5

### arcee-ai/trinity-large-thinking

Trinity-Large-Thinking is a reasoning-optimized variant of Arcee AI's Trinity-Large family — a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token. Built on Trinity-Large-Base and post-trained with extended chain-of-thought reasoning and agentic RL, Trinity-Large-Thinking delivers state-of-the-art performance on agentic benchmarks while maintaining strong general capabilities.

- **Max output tokens:** 80K
- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/trinity-large-thinking

### baseten/elephant-1

Elephant-1 is a 9B decision model from Baseten's Base Labs for classification, routing, and scoring. It evaluates structured context against choice, yes/no, and ordered scoring questions and returns answer probabilities and confidence in a single forward pass, without generating free-form text. Applications can use these probabilities to select answers or set confidence thresholds.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/elephant-1

### baseten/mammoth-1

Mammoth-1 is a 27B decision model from Baseten's Base Labs, fine-tuned from Qwen3.5-27B for classification, routing, and scoring. It answers multiple-choice, yes/no, and ordered scoring questions with probabilities over predefined options rather than generated text. Shared-prefix caching reuses context across questions, while per-option probabilities support confidence thresholds and escalation decisions.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mammoth-1

### bfl/flux-3-video

FLUX 3 is a new multimodal foundation model that learns from images, videos, and audio within a unified architecture.

- **Details:** https://vercel.com/ai-gateway/models/flux-3-video

### bfl/flux-3-image

FLUX 3 Image is Black Forest Labs’ image generation and editing model. It supports text-to-image generation, editing with up to 10 reference images, and layouts specified with bounding boxes, across five resolution classes from 768 × 768 to 4K.

- **Image cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-3-image

### bfl/flux-3-image-klein

FLUX 3 Image Klein is Black Forest Labs’ smaller, faster image generation and editing model, with 5.6 billion parameters and open weights. It supports text-to-image generation, image editing, multiple reference images, and layouts specified with bounding boxes. The hosted API produces images across five resolution classes from 768 × 768 to 4K.

- **Image cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-3-image-klein

### bfl/flux-pro-1.0-fill

A state-of-the-art inpainting model, enabling editing and expansion of real and generated images given a text description and a binary mask.
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Image cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-pro-1.0-fill

### bfl/flux-kontext-max

FLUX.1 Kontext creates images from text prompts with unique capabilities for character consistency and advanced editing. It also edits images using simple text prompts. No complex workflows or fine-tuning needed.
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Image cost:** $0.08/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-kontext-max

### bfl/flux-kontext-pro

FLUX.1 Kontext creates images from text prompts with unique capabilities for character consistency and advanced editing. It also edits images using simple text prompts. No complex workflows or fine-tuning needed.
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-kontext-pro

### bfl/flux-2-flex

FLUX.2 is a completely new base model trained for visual intelligence, not just pixel generation, setting a new standard for both image generation and image editing. With FLUX.2 models you can expect the highest quality, higher resolutions (up to 4MP), and new capabilities like multi-ref images. FLUX.2 [flex] supports customizable image generation and editing with adjustable steps and guidance. It's better at typography and text rendering. It supports up to 10 reference images (up to 14 MP total input).
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Details:** https://vercel.com/ai-gateway/models/flux-2-flex

### bfl/flux-2-klein-4b

FLUX.2 [klein] is Black Forest Labs' fastest image model yet - it unifies image generation and editing in a single, compact model. Delivering state-of-the-art quality with end-to-end inference in less than a second. Enabling interactive workflows, real-time previews, and latency-critical applications.

- **Details:** https://vercel.com/ai-gateway/models/flux-2-klein-4b

### bfl/flux-2-klein-9b

FLUX.2 [klein] is Black Forest Labs' fastest image model yet - it unifies image generation and editing in a single, compact model. Delivering state-of-the-art quality with end-to-end inference in less than a second. Enabling interactive workflows, real-time previews, and latency-critical applications.

- **Details:** https://vercel.com/ai-gateway/models/flux-2-klein-9b

### bfl/flux-2-max

FLUX.2 [max] offers image generation and image editing with the highest quality available. It delivers state-of-the-art image generation and advanced image editing with exceptional realism, precision, and consistency. Built for professional use, FLUX.2 [max] produces production-ready outputs for marketing teams, creatives, filmmakers, and creators around the world.

- **Max output tokens:** 67.3K
- **Details:** https://vercel.com/ai-gateway/models/flux-2-max

### bfl/flux-2-pro

FLUX.2 is a completely new base model trained for visual intelligence, not just pixel generation, setting a new standard for both image generation and image editing. With FLUX.2 models you can expect the highest quality, higher resolutions (up to 4MP), and new capabilities like multi-ref images. FLUX.2 [pro] supports generation, editing, and multiple reference images (up to 9 MP total input).
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Max output tokens:** 67.3K
- **Details:** https://vercel.com/ai-gateway/models/flux-2-pro

### bfl/flux-pro-1.1

FLUX1.1 [pro] is the standard for text-to-image generation with fast, reliable and consistently stunning results.
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-pro-1.1

### bfl/flux-pro-1.1-ultra

FLUX1.1 [pro] Ultra delivers ultra-fast, ultra high-resolution image creation - with more pixels in every picture. Generate varying aspect ratios from text, at 4MP resolution fast.
This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).

- **Image cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/flux-pro-1.1-ultra

### bytedance/seed-1.8

Bytedance Seed 1.8 features stronger multimodal understanding and agent capabilities. The model delivers superior performance across a wide range of complex real-world tasks, helping enterprises create greater value.

- **Max output tokens:** 64K
- **Cached input cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/seed-1.8

### bytedance/seed-1.6

ByteDance's new multimodal deep-thinking model, supporting both text and visual inputs with enhanced reasoning capabilities.

- **Max output tokens:** 32K
- **Cached input cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/seed-1.6

### bytedance/seed-2.1-turbo

Seed 2.1 Turbo is ByteDance’s multimodal language model supporting text, image, and video inputs with text output. It offers a 262,144-token context window, reasoning, function calling, and structured JSON output.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.10/1M
- **Details:** https://vercel.com/ai-gateway/models/seed-2.1-turbo

### bytedance/seedance-2.0

Built with a unified multimodal audio-video joint generation architecture, Seedance 2.0 supports four input modalities: text, image, audio, and video. Compared with Version 1.5, Seedance 2.0 delivers a substantial leap in generation quality. It achieves a higher usability rate for complex interaction and motion scenes, with significant improvements in physical accuracy, visual realism, and controllability, making it well-suited for high-quality creation scenarios.

- **Details:** https://vercel.com/ai-gateway/models/seedance-2.0

### bytedance/seedance-2.0-fast

Seedance 2.0 Fast is a new-generation multimodal video creation model, inheriting the core functions and advantages of Seedance 2.0, with faster speed.

- **Details:** https://vercel.com/ai-gateway/models/seedance-2.0-fast

### bytedance/seedance-2.0-mini

Seedance 2.0 Mini is a cost-effective multimodal video creation model, retaining the core functions and advantages of Seedance 2.0 at a lower price.

- **Details:** https://vercel.com/ai-gateway/models/seedance-2.0-mini

### bytedance/seedance-2.5

Seedance 2.5 is a next-generation audio-video joint generation model, built for 30-second storytelling with precise reference control and powerful editing capabilities.

- **Details:** https://vercel.com/ai-gateway/models/seedance-2.5

### bytedance/seedance-v1.0-pro

A video generation model that supports multi-shot storytelling. It excels in semantic understanding and instruction following, producing smooth, detailed, and cinematic 1080P HD videos.

- **Details:** https://vercel.com/ai-gateway/models/seedance-v1.0-pro

### bytedance/seedance-v1.0-pro-fast

Seedance 1.0 Pro Fast delivers top performance at an unbeatable price, balancing quality, speed, and cost. Built on Seedance 1.0 Pro’s core strengths, it’s faster and more cost-efficient for creators.

- **Details:** https://vercel.com/ai-gateway/models/seedance-v1.0-pro-fast

### bytedance/seedance-v1.5-pro

ByteDance's Seedance 1.5 Pro is a professional video model using V2A native generation for integrated, synced audio-visual output, enhancing efficiency of professional video creation.

- **Details:** https://vercel.com/ai-gateway/models/seedance-v1.5-pro

### bytedance/seedream-4.0

Seedream 4.0 is a SOTA multimodal image creation model built on leading architecture. It breaks through the boundaries of traditional text-to-image models by natively supporting text, single-image, and multi-image inputs. Users can freely combine text and images to achieve diverse creative modes within a single model—such as multi-image blending, image editing, and sequentially batch image generation, featuring subject consistency, making image creation more free and controllable.

- **Image cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/seedream-4.0

### bytedance/seedream-4.5

Seedream 4.5 is the latest in-house image generation model developed by ByteDance. Compared with Seedream 4.0, it delivers comprehensive improvements—especially in editing consistency, including better preservation of subject details, lighting, and color tone. It also enhances portrait refinement and small-text rendering. The model’s multi-image composition capabilities have been significantly strengthened, and both reasoning performance and visual aesthetics continue to advance, enabling more accurate and artistically expressive image generation.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/seedream-4.5

### bytedance/seedream-5.0-lite

ByteDance-Seedream-5.0-lite is the latest image generation model released by BytePlus. For the first time, it introduces web-connected retrieval, enabling the model to fuse real-time online information to significantly improve the timeliness and relevance of generated images. The model’s reasoning and comprehension capabilities are further upgraded, allowing it to accurately interpret complex prompts and visual inputs. In addition, ByteDance-Seedream-5.0-lite delivers notable improvements in global knowledge coverage, reference consistency, and professional-grade scene generation, making it well suited for enterprise-level visual creation workflows.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/seedream-5.0-lite

### bytedance/seedream-5.0-pro

Seedream-5.0-Pro, ByteDance's newest image generation model, delivers comprehensive upgrades for complex, lifelike image creation and editing, ushering in a new phase of controllable visual production. It stands out with precise editing control, robust commercial applicability and natural rendering results.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/seedream-5.0-pro

### callstack/apex

Apex is Callstack’s agentic coding model for mobile and web development, optimized for React Native, React, and Next.js. With a 256K context window, it supports feature development, debugging, code reviews, testing, and large refactors.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.20/1M
- **Details:** https://vercel.com/ai-gateway/models/apex

### cohere/rerank-v3.5

A model that allows for re-ranking English Language documents and semi-structured data (JSON).

- **Max output tokens:** 4.1K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/rerank-v3.5

### cohere/rerank-v4-fast

A light version of Rerank 4 Pro, this is a multilingual model that allows for re-ranking English and non-english documents and semi-structured data (JSON). This model is better suited for low latency and high throughput use-cases than its pro variant.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-v4-fast

### cohere/rerank-v4-pro

A multilingual model that allows for re-ranking English and non-english documents and semi-structured data (JSON). This model is better suited for state-of-the-art quality and complex use-cases than its fast variant.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-v4-pro

### cohere/rerank-v5-fast

Cohere Rerank 5 Fast is a multilingual text reranking model optimized for low latency and high throughput in search and agent workflows. It ranks documents and semi-structured data across more than 100 languages, with calibrated relevance scores and a 32,768-token context window per query-document pair.

- **Max output tokens:** 32.8K
- **Details:** https://vercel.com/ai-gateway/models/rerank-v5-fast

### cohere/rerank-v5-pro

Cohere Rerank 5 Pro is a multilingual text reranking model for complex queries and long or domain-specific documents. It ranks documents and semi-structured data across more than 100 languages, with calibrated relevance scores and a 32,768-token context window per query-document pair.

- **Max output tokens:** 32.8K
- **Details:** https://vercel.com/ai-gateway/models/rerank-v5-pro

### cohere/command-a

Command A is Cohere's most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024.

- **Max output tokens:** 8K
- **Details:** https://vercel.com/ai-gateway/models/command-a

### cohere/embed-v5.0-fast

Cohere's lightweight embedding model optimized for low latency and high throughput in interactive search, agent loops, and high-volume retrieval. Embed 5 Fast supports text, images, and mixed content, over 100 languages, and a 128,000-token context window, with output dimensions from 256 to 2,048. It shares an embedding space with Embed 5 Pro.

- **Details:** https://vercel.com/ai-gateway/models/embed-v5.0-fast

### cohere/embed-v5.0-pro

Cohere's most capable embedding model for quality-critical enterprise retrieval across text, images, and mixed content such as PDFs. Embed 5 Pro supports over 100 languages and a 128,000-token context window, with output dimensions from 256 to 2,048. It shares an embedding space with Embed 5 Fast, enabling Pro indexing with Fast querying.

- **Details:** https://vercel.com/ai-gateway/models/embed-v5.0-pro

### cohere/embed-v4.0

A model that allows for text, images, or mixed content to be classified or turned into embeddings.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/embed-v4.0

### convaiinnovations/laya

Laya is Convai Innovations’ System One evaluation model for fast, structured decisions. It evaluates shared state against choice, score, and yes/no questions and returns typed answers with probabilities in a single forward pass. It supports classification, routing, guardrail checks, and rubric-based assessment without generating free-form text.

- **Details:** https://vercel.com/ai-gateway/models/laya

### deepseek/deepseek-r1

The DeepSeek R1 model has undergone a minor version upgrade, with the current version being DeepSeek-R1-0528.

- **Max output tokens:** 16.4K
- **Cached input cost:** $0.35/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-r1

### deepseek/deepseek-v3.1

DeepSeek-V3.1 is post-trained on the top of DeepSeek-V3.1-Base, which is built upon the original V3 base checkpoint through a two-phase long context extension approach, following the methodology outlined in the original DeepSeek-V3 report. We have expanded our dataset by collecting additional long documents and substantially extending both training phases. The 32K extension phase has been increased 10-fold to 630B tokens, while the 128K extension phase has been extended by 3.3x to 209B tokens. Additionally, DeepSeek-V3.1 is trained using the UE8M0 FP8 scale data format to ensure compatibility with microscaling data formats.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v3.1

### deepseek/deepseek-v3.2

DeepSeek-V3.2: Official successor to V3.2-Exp.

- **Max output tokens:** 8K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v3.2

### deepseek/deepseek-v3.2-thinking



- **Max output tokens:** 8K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v3.2-thinking

### deepseek/deepseek-v4-flash

DeepSeek-V4 series incorporate several key upgrades in architecture and optimization: (1) a hybrid attention architecture that combines Compressed Sparse Attention (CSA)
and Heavily Compressed Attention (HCA) to improve long-context efficiency; (2) ManifoldConstrained Hyper-Connections (mHC) that enhance conventional residual connections; (3)
and the Muon optimizer for faster convergence and greater training stability

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4-flash

### deepseek/deepseek-v4-flash-0731

DeepSeek-V4 series incorporate several key upgrades in architecture and optimization: (1) a hybrid attention architecture that combines Compressed Sparse Attention (CSA)
and Heavily Compressed Attention (HCA) to improve long-context efficiency; (2) ManifoldConstrained Hyper-Connections (mHC) that enhance conventional residual connections; (3)
and the Muon optimizer for faster convergence and greater training stability

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4-flash-0731

### deepseek/deepseek-v4-flash-vision-exp

DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of DeepSeek V4 Flash, adding image understanding while retaining its text, agent, reasoning, and world-knowledge capabilities. It is suited for document and chart understanding, visual question answering, and multimodal agent workflows.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4-flash-vision-exp

### deepseek/deepseek-v4-pro

DeepSeek-V4 series incorporate several key upgrades in architecture and optimization: (1) a hybrid attention architecture that combines Compressed Sparse Attention (CSA)
and Heavily Compressed Attention (HCA) to improve long-context efficiency; (2) ManifoldConstrained Hyper-Connections (mHC) that enhance conventional residual connections; (3)
and the Muon optimizer for faster convergence and greater training stability

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4-pro

### deepseek/deepseek-v4-pro-0813

This is the 8/13 updated weights version of DeepSeek V4 Pro.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4-pro-0813

### deepseek/deepseek-v4.1-flash

DeepSeek V4.1 Flash is an AI model from DeepSeek designed for fast, efficient multimodal interactions. It combines native multimodal capabilities with a new architecture focused on stronger performance, faster responses, and lower computational cost.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.00/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4.1-flash

### deepseek/deepseek-v4.1-flash-1017

10/17 snapshot of DeepSeek V4.1 Flash

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.00/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4.1-flash-1017

### deepseek/deepseek-v4.1-flash-1017-fast

Higher-throughput variant of DeepSeek V4.1 Flash 1017 served by Baseten for multimodal reasoning, tool calling, and structured output, with automatic prefix caching.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4.1-flash-1017-fast

### deepseek/deepseek-v4.1-flash-fast

Higher-throughput variant of DeepSeek V4.1 Flash for responsive coding, reasoning, and agentic workloads.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/deepseek-v4.1-flash-fast

### elevenlabs/eleven-flash-v2

English-only text-to-speech optimized for ultra-low latency in real-time applications. Accepts up to 30,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-flash-v2

### elevenlabs/eleven-flash-v2.5

Ultra-low-latency text-to-speech for real-time applications in 32 languages. Accepts up to 40,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-flash-v2.5

### elevenlabs/eleven-multilingual-v2

Lifelike text-to-speech with consistent voice quality for long-form audio in 29 languages. Accepts up to 10,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-multilingual-v2

### elevenlabs/music-v2

Music generation from text prompts, billed per second of generated audio.

- **Details:** https://vercel.com/ai-gateway/models/music-v2

### elevenlabs/music-v2.5

Music generation from text prompts, billed per second of generated audio.

- **Details:** https://vercel.com/ai-gateway/models/music-v2.5

### elevenlabs/eleven-realtime-v1-max

A speech-to-speech voice agent optimized for complex instructions, reasoning, and multi-step tool use, with speech recognition, turn-taking, and voice generation tuned as one system. Supports more than 90 languages through an OpenAI-compatible Realtime WebSocket API. Billed for connected session duration plus input and output tokens.

- **Cached input cost:** $0.09/1M
- **Details:** https://vercel.com/ai-gateway/models/eleven-realtime-v1-max

### elevenlabs/eleven-realtime-v1-mini

A speech-to-speech voice agent optimized for fast, natural conversation, with speech recognition, turn-taking, interruption handling, and voice generation tuned as one system. Supports more than 90 languages through an OpenAI-compatible Realtime WebSocket API. Billed for connected session duration plus input and output tokens.

- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/eleven-realtime-v1-mini

### elevenlabs/sound-effects-v2

Sound effects generation from text prompts, billed per second of generated audio.

- **Details:** https://vercel.com/ai-gateway/models/sound-effects-v2

### elevenlabs/eleven-v3

Emotionally rich text-to-speech with inline audio tags for delivery and expression, supporting more than 70 languages. Accepts up to 5,000 characters per request and is billed per input character, including audio tags.

- **Details:** https://vercel.com/ai-gateway/models/eleven-v3

### elevenlabs/eleven-v3-conversational

A low-latency Eleven v3 variant tuned for conversational agents, with expressive audio tags and support for more than 70 languages. Accepts up to 5,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-v3-conversational

### elevenlabs/eleven-v4

ElevenLabs' expressive text-to-speech model with audio tags for emotion and delivery, multilingual speech, and multi-speaker dialogue support. Accepts up to 10,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-v4

### elevenlabs/eleven-v4-turbo

A low-latency Eleven v4 variant for real-time and conversational applications, with expressive audio tags and multilingual speech. Accepts up to 10,000 characters per request and is billed per input character.

- **Details:** https://vercel.com/ai-gateway/models/eleven-v4-turbo

### elevenlabs/scribe-v2

Multilingual speech-to-text with speaker diarization, word-level timestamps, and audio event tagging. Supports more than 90 languages and is billed per second of input audio.

- **Details:** https://vercel.com/ai-gateway/models/scribe-v2

### elevenlabs/scribe-v2-realtime

Real-time speech-to-text over WebSocket with manual or voice-activity-based commits, optional word timestamps, and language detection. Supports PCM and mu-law audio and is billed per second of streamed audio.

- **Details:** https://vercel.com/ai-gateway/models/scribe-v2-realtime

### elorian/glimpse

Glimpse is Elorian's visual reasoning model for image and video understanding, with a 220,000-token context window. It tracks objects across motion and viewpoints, interprets spatial relationships, and reasons about physical cause and effect. Designed for scientific chart analysis, video captioning, robotics planning, and visual counting, it combines multimodal understanding with an agent trained to use code and visual inspection tools.

- **Max output tokens:** 4.1K
- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/glimpse

### fireworks/ember-1

Ember-1 is Fireworks’ research-preview reasoning model built on Kimi K3 for coding and agentic workflows. It supports image input, function calling, and a 1M-token context window. Fireworks reports approximately 40% fewer generated tokens while maintaining comparable quality across its evaluations, helping reduce the cost of workloads with extensive reasoning.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.30/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/ember-1

### fish-audio/drama-3

Drama 3 is Fish Audio’s text-to-speech model, currently in preview, that lets you direct voice performances using natural-language instructions. It supports control over emotion, pacing, pauses, and vocal sounds, with tags that apply effects to specific words. Multi-speaker dialogue and directions written in different languages make it well suited to expressive voiceovers, character dialogue, and audio storytelling.

- **Details:** https://vercel.com/ai-gateway/models/drama-3

### fish-audio/s1

Fish Audio S1 is trained on over 2 million hours of audio with online RLHF (GRPO). It achieves 0.8% WER and 0.4% CER on Seed TTS Eval. S1 supports open-domain emotion, tone, and special effect markers.

- **Details:** https://vercel.com/ai-gateway/models/s1

### fish-audio/s2-pro

Fish Audio S2 is the second-generation TTS model from Fish Audio. It's trained on over 10 million hours of audio across approximately 80 languages, and it introduces inline tag control: natural-language instructions embedded directly in your script at any position, giving you fine-grained direction over how speech is delivered at the word or phrase level.

- **Details:** https://vercel.com/ai-gateway/models/s2-pro

### fish-audio/s2.1-pro

S2.1 Pro is Fish Audio's current state-of-the-art voice model — the best model we have, now available to every developer for free via API. It is a neural speech synthesis model designed for production-grade AI voice generation, with particular strengths in low-latency streaming, multilingual TTS, and voice cloning. 

- **Details:** https://vercel.com/ai-gateway/models/s2.1-pro

### fish-audio/transcribe-1

Turn spoken audio into accurate text — with timed segments — using Fish Audio’s ASR model. Send an audio file, get back the transcript, its duration, and timestamped segments.

- **Details:** https://vercel.com/ai-gateway/models/transcribe-1

### google/gemini-2.5-flash

Gemini 2.5 Flash is a thinking model that offers great, well-rounded capabilities. It is designed to offer a balance between price and performance with multimodal support and a 1M token context window.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-2.5-flash

### google/gemini-2.5-flash-lite

Gemini 2.5 Flash-Lite is a balanced, low-latency model with configurable thinking budgets and tool connectivity (e.g., Google Search grounding and code execution). It supports multimodal input and offers a 1M-token context window.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-2.5-flash-lite

### google/gemini-2.5-pro

Gemini 2.5 Pro is our most advanced reasoning Gemini model, capable of solving complex problems. Gemini 2.5 Pro can comprehend vast datasets and challenging problems from different information sources, including text, audio, images, video, and even entire code repositories.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-2.5-pro

### google/gemini-3-flash

Google's most intelligent model built for speed, combining frontier intelligence with superior search and grounding.

- **Max output tokens:** 65K
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3-flash

### google/gemini-3.1-flash-image

Gemini 3.1 Flash Image is optimized for image understanding and generation and offers a balance of price and performance.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.1-flash-image

### google/gemini-3.1-flash-image-preview

Gemini 3.1 Flash Image (Nano Banana 2) is optimized for image understanding and generation and offers a balance of price and performance.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.1-flash-image-preview

### google/gemini-3.1-flash-lite

Gemini 3.1 Flash Lite outperforms 2.5 Flash Lite on overall quality and lands close to 2.5 Flash performance across key capability areas. It is a workhorse model for high-volume use cases, with improvements across audio input/ASR, RAG snippet ranking, translation, data extraction, and code completion.

- **Max output tokens:** 65K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.1-flash-lite

### google/gemini-3.1-flash-lite-image

Gemini 3.1 Flash-Lite Image (Nano Banana 2 Lite) is Google's fastest image generation model enabling rapid creation and iteration.

- **Max output tokens:** 4.1K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.1-flash-lite-image

### google/gemini-3.1-pro-preview

Improved SWE and agentic capabilities, improved token efficiency and thinking, and expanded thinking levels.

- **Max output tokens:** 64K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.1-pro-preview

### google/gemini-3.5-flash

Google's latest model, highly optimized for coding proficiency and parallel agentic execution loops. Defaults to medium thinking effort for faster and more cost-efficient responses.

- **Max output tokens:** 64K
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.5-flash

### google/gemini-3.5-flash-lite

Gemini 3.5 Flash Lite features upgraded agentic capabilities, making the model ideal for subagents in complex workflows.

- **Max output tokens:** 65K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.5-flash-lite

### google/gemini-3.5-transcribe

Transcription by google

- **Details:** https://vercel.com/ai-gateway/models/gemini-3.5-transcribe

### google/gemini-3.5-transcribe-live

Live transcription by Google

- **Details:** https://vercel.com/ai-gateway/models/gemini-3.5-transcribe-live

### google/gemini-3.6-flash

Gemini 3.6 Flash delivers higher quality across coding, agentic workflows, and web development with reduced token consumption and fewer model calls compared to previous model iterations.

- **Max output tokens:** 64K
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.6-flash

### google/gemini-3.7-flash

Gemini 3.7 Flash is the next iteration in the Gemini 3 model family, featuring algorithmic improvements to its core reasoning foundation. It supports customizable thinking configurations to control the mix of quality, cost and latency.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.7-flash

### google/gemini-3.8-flash

Gemini 3.8 Flash is the next iteration in the Gemini 3 model family, featuring algorithmic improvements to its core reasoning foundation. It supports customizable thinking configurations to control the mix of quality, cost and latency.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.8-flash

### google/gemini-3.8-flash-tts

Gemini 3.8 Flash TTS is Google’s text-to-speech model for expressive voice generation and custom character design across more than 100 languages. It creates original voices from natural language prompts and offers granular control over acting cues, pacing, accents, and conversational reactions. Long-form audio and native two-speaker dialogue make it well suited to immersive audiobooks, games, podcasts, and interactive media

- **Max output tokens:** 16.4K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.8-flash-tts

### google/gemini-3.8-flash-lite-tts

Gemini 3.8 Flash-Lite TTS is Google’s text-to-speech model optimized for high-volume, cost-efficient audio generation across more than 100 languages. It combines expressive speech with fine-grained control over tone, pacing, and line-by-line delivery, supporting long-form narration and native two-speaker conversations. Built for scale, it helps developers and creators power multilingual dubbing, audio content production, and expressive voice agents.

- **Max output tokens:** 16.4K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3.8-flash-lite-tts

### google/gemini-3.8-live

Gemini 3.8 Live is Google's model for real-time audio conversations. It supports spoken interactions for developers building voice assistants, conversational experiences, and applications that respond to users through audio.

- **Details:** https://vercel.com/ai-gateway/models/gemini-3.8-live

### google/gemini-3.8-live-extended-thinking

Gemini 3.8 Live Extended Thinking is Google's real-time audio model with extended thinking capabilities. It combines spoken interaction with additional reasoning for voice assistants and other conversational applications.

- **Details:** https://vercel.com/ai-gateway/models/gemini-3.8-live-extended-thinking

### google/gemini-4-argon

Gemini 4 Argon is Google’s multimodal reasoning model, supporting text, image, and file inputs with configurable reasoning levels. It combines tool calling, structured outputs, and Google Search and Maps grounding to support document analysis and agent workflows. Long-form generation supports continuation across requests, allowing applications to resume output beyond a single request’s token limit.

- **Max output tokens:** 1M
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-4-argon

### google/gemini-embedding-001

State-of-the-art embedding model with excellent performance across English, multilingual and code tasks.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-embedding-001

### google/gemini-embedding-2

Google’s first fully multimodal Embedding model that is capable of mapping text, image, video, audio, and PDFs and their interleaved combinations thereof into a single, unified vector space. Built on the Gemini architecture, it supports 100+ languages.

- **Details:** https://vercel.com/ai-gateway/models/gemini-embedding-2

### google/gemini-nano-banana-2.1

Gemini Nano Banana 2.1 is optimized for image understanding and generation and offers a balance of price and performance.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-nano-banana-2.1

### google/gemini-omni-flash-preview

Gemini Omni Flash (Preview) is a multimodal model designed for video, image, and text tasks. It is optimized for video generation, offering video output alongside text responses in a single model.

- **Max output tokens:** 57.9K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-omni-flash-preview

### google/gemma-4-26b-a4b-it

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on small models) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemma-4-26b-a4b-it

### google/gemma-4-31b-it

Gemma 4 31B is engineered to tackle the most demanding enterprise workloads and complex reasoning tasks. With an expansive 256K-token context window, the 31B model can effortlessly ingest entire codebases, and massive sets of images in a single prompt.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.06/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemma-4-31b-it

### google/gemini-2.5-flash-image

Nano Banana (Gemini 2.5 Flash Image) is Google's first fully hybrid reasoning model, letting developers turn thinking on or off and set thinking budgets to balance quality, cost, and latency. Upgraded for rapid creative workflows, it can generate interleaved text and images and supports conversational, multi‑turn image editing in natural language. It’s also locale‑aware, enabling culturally and linguistically appropriate image generation for audiences worldwide.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-2.5-flash-image

### google/gemini-3-pro-image

Nano Banana Pro (Gemini 3 Pro Image) builds on Nano Banana's generation capabilities into a new era of studio-quality, functional design to help you create and edit high-fidelity, production-ready visuals with unparalleled precision and control. Improvements include enhanced world knowledge and reasoning, dynamic text and translation, and studio level controls.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gemini-3-pro-image

### google/text-embedding-005

English-focused text embedding model optimized for code and English language tasks.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/text-embedding-005

### google/text-multilingual-embedding-002

Multilingual text embedding model optimized for cross-lingual tasks across many languages.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/text-multilingual-embedding-002

### google/veo-3.0-generate-001

Veo 3 is designed to handle a range of video generation tasks, from cinematic narratives to dynamic character animations. With Veo 3, you can create more immersive experiences by not only generating stunning visuals, but also audio like dialogue and sound effects.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/veo-3.0-generate-001

### google/veo-3.0-fast-generate-001

Veo 3 Fast is a quicker and more cost effective version of Veo 3, allowing developers to create videos with sound while maintaining high quality and optimizing for speed and business use cases. Veo 3 Fast offers both text-to-video and image-to-video modalities.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/veo-3.0-fast-generate-001

### google/veo-3.1-generate-001

Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p, 1080p or 4k videos featuring stunning realism and natively generated audio.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/veo-3.1-generate-001

### google/veo-3.1-fast-generate-001

Veo 3.1 Fast is a specialized, high-speed variant of Google DeepMind’s Veo 3.1 text-to-video model, optimized for rapid generation of 8-second, high-fidelity videos. It is designed to create cinematic, 1080p, or 720p content with improved prompt adherence and native audio, making it ideal for creating quick, high-quality video clips, social media content, and ad creatives.

- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/veo-3.1-fast-generate-001

### google/veo-3.1-lite-generate-001

Veo 3.1 Lite Preview is a high-efficiency, developer-first video model providing high-fidelity video generation, editing, and cinematic control. It leverages the state-of-the-art Veo 3.1 model to democratize professional-grade video AI by offering a scalable, programmable interface for creators and enterprises.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/veo-3.1-lite-generate-001

### heygen/heygen-video-1

HeyGen Video generates 5–15 second clips with synchronized dialogue, ambience, and sound effects from text, a first-frame image, or image and video references. It supports 480p, 768p, 1080p, and 2K output. Reference mode charges for input video seconds plus output video seconds; image and audio references have no additional charge.

- **Details:** https://vercel.com/ai-gateway/models/heygen-video-1

### inception/mercury-2

A diffusion-based reasoning LLM that generates text via parallel refinement (not token-by-token), delivering real-time latency with ~1k tokens/sec plus 128K context and built-in tool/JSON support.

- **Max output tokens:** 128K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/mercury-2

### inception/mercury-2.5

Mercury 2.5 is Inception’s diffusion-based reasoning model for chat, agents, and structured workflows, with tool calling, structured outputs, and a 260K context window.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/mercury-2.5

### inception/mercury-3

Mercury 3 is Inception’s most capable model for agentic workflows and real-time applications. It is one of the world’s fastest models, powered by diffusion.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/mercury-3

### inception/mercury-coder-small

Mercury Coder Small is ideal for code generation, debugging, and refactoring tasks with minimal latency.

- **Max output tokens:** 16.4K
- **Details:** https://vercel.com/ai-gateway/models/mercury-coder-small

### inclusionai/ling-3.0-flash

Ling 3.0 Flash is designed with token efficiency and production-scale agentic inference as key priorities, enabling developers to complete more useful work within constrained token, latency, and serving-cost budgets.

- **Max output tokens:** 32K
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/ling-3.0-flash

### inclusionai/ling-3.0-flash-fin

Ling 3.0 Flash Fin is InclusionAI’s finance-enhanced MoE language model, combining 124 billion total parameters with approximately 5.1 billion active parameters for efficient financial reasoning. Its 256K context window, function calling, and support for complex, multi-step investment workflows make it ideal for financial research, analysis, long-horizon planning, and execution, while retaining strong capabilities in coding and mathematics.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/ling-3.0-flash-fin

### inclusionai/ling-3.0-flash-sante

Ling-3.0-Flash-Sante is inclusionAI’s language model specialized for health and medicine, built on a Mixture-of-Experts architecture with 124 billion total parameters and approximately 5.1 billion active per token. With a 256K context window and function calling, it supports medical knowledge reasoning, evidence-based retrieval, and complex medical workflows while retaining general reasoning, coding, and agentic capabilities.

- **Max output tokens:** 32K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/ling-3.0-flash-sante

### inclusionai/ling-3.0-flash-vl

Ling 3.0 Flash VL builds on Ling 3.0 Flash with stronger language capabilities, native visual perception, and visual agent capabilities. It supports text, image, and video inputs with text output, reasoning, and function calling.

- **Max output tokens:** 32K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/ling-3.0-flash-vl

### inclusionai/ling-3.1-flash

Ling 3.1 Flash is InclusionAI's hybrid reasoning language model for coding, multi-step analysis, and tool-using agents. It has 560B total parameters with 25B activated parameters and supports long-context text workflows.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/ling-3.1-flash

### inference-net/schematron-v2-small

Schematron V2 Small is a specialized model from Inference.net for extracting structured data from HTML. It uses a supplied JSON Schema to produce conforming JSON and is designed for complex schemas and long pages.

- **Max output tokens:** 4.1K
- **Cached input cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/schematron-v2-small

### inference-net/schematron-v2-turbo

Schematron V2 Turbo is a specialized model from Inference.net for extracting structured data from HTML. It uses a supplied JSON Schema to produce conforming JSON and is optimized for high-throughput extraction.

- **Max output tokens:** 8.2K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/schematron-v2-turbo

### interfaze/interfaze-beta

Interfaze is an AI model built on a new architecture that merges specialized DNN/CNN models with LLMs for developer tasks that require deterministic output and high consistency like OCR, scraping, classification, STT and more.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/interfaze-beta

### klingai/kling-v2.5-turbo-i2v

Kling 2.5 Turbo is a major update to the AI video generation model focused on significantly improving speed, video quality, temporal stability, and creative control for creators, making professional-grade AI-generated video faster, more coherent, and easier to direct from text prompts.

- **Details:** https://vercel.com/ai-gateway/models/kling-v2.5-turbo-i2v

### klingai/kling-v2.5-turbo-t2v

Kling 2.5 Turbo is a major update to the AI video generation model focused on significantly improving speed, video quality, temporal stability, and creative control for creators, making professional-grade AI-generated video faster, more coherent, and easier to direct from text prompts.

- **Details:** https://vercel.com/ai-gateway/models/kling-v2.5-turbo-t2v

### klingai/kling-v2.6-i2v

Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience. 

- **Details:** https://vercel.com/ai-gateway/models/kling-v2.6-i2v

### klingai/kling-v2.6-motion-control

Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience. 

- **Details:** https://vercel.com/ai-gateway/models/kling-v2.6-motion-control

### klingai/kling-v2.6-t2v

Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience. 

- **Details:** https://vercel.com/ai-gateway/models/kling-v2.6-t2v

### klingai/kling-v3.0-i2v

Build upon an All-in-One product framework, the Kling 3.0 model series supports full multimodal input and output spanning text, images, audio, and video, bringing the understanding, generation, and editing of video together in one streamlined AI workflow. The models integrate multiple tasks, including text-to-video, image-to-video, reference-to-video, and in-video editing, into a single, native multimodal architecture, enabling the models to follow complex narrative logic, deliver precise shot control, and maintain strong prompt adherence.

- **Details:** https://vercel.com/ai-gateway/models/kling-v3.0-i2v

### klingai/kling-v3.0-motion-control

Kling 3.0 delivers a major leap in character fidelity for motion-driven generation, with stable facial features across multi-angle and long-duration motion, accurate complex emotions from multi-image face references, identity preservation through partial occlusions (hats, hands, fans), and steady clarity as the camera zooms, pans, or tracks.

- **Details:** https://vercel.com/ai-gateway/models/kling-v3.0-motion-control

### klingai/kling-v3.0-t2v

Build upon an All-in-One product framework, the Kling 3.0 model series supports full multimodal input and output spanning text, images, audio, and video, bringing the understanding, generation, and editing of video together in one streamlined AI workflow. The models integrate multiple tasks, including text-to-video, image-to-video, reference-to-video, and in-video editing, into a single, native multimodal architecture, enabling the models to follow complex narrative logic, deliver precise shot control, and maintain strong prompt adherence.

- **Details:** https://vercel.com/ai-gateway/models/kling-v3.0-t2v

### liquid/lfm2.5-24b-a2b

LFM2.5-24B-A2B is a text-generation model from Liquid AI with a 128,000-token context window.

- **Max output tokens:** 8.2K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/lfm2.5-24b-a2b

### liquid/d1

Liquid d1 is a decision model for classification, routing, and scoring. It evaluates shared state against typed Choice, Score, and yes/no questions and returns structured answers with calibrated probabilities through the TypeSafe-compatible System One API.

- **Details:** https://vercel.com/ai-gateway/models/d1

### liquid/d1-ultra

Liquid d1-ultra is a multimodal decision model for classification, routing, and scoring across text and images, with a 32,768-token context window. It returns calibrated probabilities for yes/no, multiple-choice, and scale-based questions in a single forward pass without generating output tokens. It is designed for demanding real-world tasks such as visual defect inspection and evaluating damage claims against customer photos.

- **Details:** https://vercel.com/ai-gateway/models/d1-ultra

### meituan/longcat-2.5-preview

LongCat-2.5-Preview is Meituan’s multimodal reasoning model for coding and agentic workflows, with a 1M-token context window and support for text and image inputs. It combines image understanding, visual question answering, and content summarization with code generation, code understanding, and automated programming, and supports tool calling and optional thinking for complex tasks.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/longcat-2.5-preview

### meta/llama-3.1-70b

An update to Meta Llama 3 70B Instruct that includes an expanded 128K context length, multilinguality and improved reasoning capabilities.

- **Max output tokens:** 8.2K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/llama-3.1-70b

### meta/llama-3.1-8b

An update to Meta Llama 3 8B Instruct that includes an expanded 128K context length, multilinguality and improved reasoning capabilities.

- **Max output tokens:** 16.4K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/llama-3.1-8b

### meta/llama-3.3-70b

Where performance meets efficiency. This model supports high-performance conversational AI designed for content creation, enterprise applications, and research, offering advanced language understanding capabilities, including text summarization, classification, sentiment analysis, and code generation.

- **Max output tokens:** 8.2K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/llama-3.3-70b

### meta/llama-4-maverick

As a general purpose LLM, Llama 4 Maverick contains 17 billion active parameters, 128 experts, and 400 billion total parameters, offering high quality at a lower price compared to Llama 3.3 70B.

- **Max output tokens:** 8.2K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/llama-4-maverick

### meta/llama-4-scout

The Llama 4 collection of models are natively multimodal AI models that enable text and multimodal experiences. These models leverage a mixture-of-experts architecture to offer industry-leading performance in text and image understanding. Llama 4 Scout, a 17 billion parameter model with 16 experts. Served by DeepInfra.

- **Max output tokens:** 8.2K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/llama-4-scout

### meta/muse-glimmer-30b



- **Max output tokens:** 131.1K
- **Cached input cost:** $0.04/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-glimmer-30b

### meta/muse-image-1.0

Muse Image is the first image generation model from Meta Superintelligence Labs, it uses advanced reasoning to understand complex prompts, seamlessly blending multiple photos into high-quality creations you can download and share anywhere.


- **Image cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-image-1.0

### meta/muse-spark-1.1

Muse Spark 1.1 is strongest at agentic performance, tool use, and computer use. It does well on long-running tasks with 1M token context window, can delegate execution to sub-agents running in parallel, and is trained to use computer interfaces on desktop, mobile, or browser.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-1.1

### meta/muse-spark-1.2

A coding-optimized model purpose-built for agentic workflows. Improvements to code generation, debugging, and codebase understanding — with a 1M context window that handles your entire project in one session.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-1.2

### meta/muse-spark-1.2-contributor

A coding-optimized model with pricing designed for builders. Same model, same capabilities — up to 95% less than Standard. Your inputs and outputs are used to train and improve Meta's AI models.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-1.2-contributor

### meta/muse-spark-1.3

Muse Spark 1.3 is Meta’s multimodal reasoning model for long-horizon agentic and coding workflows. With a 1M-token context window, reliable tool calling, higher first-attempt accuracy, and native understanding of video, images, and documents, it helps developers build capable coding agents and AI development workflows with fewer unnecessary turns and cleaner output.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-1.3

### meta/muse-spark-1.3-contributor

Muse Spark 1.3 Contributor is Meta’s cost-optimized API option for agentic coding workflows. It offers a 1M-token context window, dependable tool calling, and multimodal perception at $0.10 per million input tokens and $0.20 per million output tokens, with usage permitted to improve Meta’s products.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-1.3-contributor

### meta/muse-spark-open

Muse Spark Open is Meta’s open-weight reasoning model for coding and agentic workflows. Its sparse mixture-of-experts architecture has 816 billion total parameters with approximately 42 billion active per token, combining code generation, tool use, and image understanding. A one-million-token context window supports repository analysis, multi-file changes, debugging, and long-running agent tasks.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-spark-open

### meta/muse-wonder-1.0

Muse Wonder is Meta’s model for frontier coding, long-running agents, and advanced computer use. It plans long tasks, delegates work to sub-agents, and recovers from errors.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/muse-wonder-1.0

### microsoft/microsoft-decision-1

Microsoft Decision-1 is a text decision model from Microsoft AI for classification, routing, and evaluation. It answers yes/no questions, selects among predefined options, and scores inputs against custom criteria, returning calibrated probabilities and confidence values. With support for multiple questions per request and a 32,768-token context window, it helps developers triage support requests, guide automated workflows, and evaluate AI responses against a rubric.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/microsoft-decision-1

### microsoft/mai-image-2.6

MAI-Image-2.6 is a diffusion-based generative model designed for both high-quality text-to-image synthesis and precise, controllable image-to-image editing. It progressively transforms random noise into a coherent image aligned with a given text prompt, leveraging a flow-matching loss to learn a continuous transformation between the noise distribution and the data distribution. MAI-Image-2.6 excels at precise, surgical edits with consistency - enabling users and developers to make targeted object edits, adapt layouts, update text, clean up artifacts like motion blur, and preserve visual consistency across iterations.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-image-2.6

### microsoft/mai-image-2.6-flash

MAI-Image-2.6 is a diffusion-based generative model designed for both high-quality text-to-image synthesis and precise, controllable image-to-image editing. It progressively transforms random noise into a coherent image aligned with a given text prompt, leveraging a flow-matching loss to learn a continuous transformation between the noise distribution and the data distribution. MAI-Image-2.6 excels at precise, surgical edits with consistency - enabling users and developers to make targeted object edits, adapt layouts, update text, clean up artifacts like motion blur, and preserve visual consistency across iterations.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-image-2.6-flash

### microsoft/mai-transcribe-2

MAI-Transcribe 2 is a multilingual speech-to-text model from Microsoft AI, ranked #1 on the FLEURS multilingual benchmark. It supports 60 languages with automatic language identification, code switching for mixed-language speech, speaker diarization, word-level timestamps, keyword biasing for domain-specific terminology, and configurable verbatim or clean transcription styles. It is suited for captions, call transcription, subtitling, accessibility, and other voice-enabled applications, and is faster than MAI-Transcribe-1.5 on long-form audio.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-transcribe-2

### microsoft/mai-transcribe-2-streaming

MAI-Transcribe-2-Streaming is a low-latency, "speech in, text out" realtime transcription API. Audio can be sent to the model as a continuous stream over a WebSocket connection and transcripts are returned incrementally as the speaker talks, with no need to wait for the utterance to finish. Intermediate results (aka Partials) update the current transcription; final results confirm a segment.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-transcribe-2-streaming

### microsoft/mai-voice-2

Microsoft AI's highest-fidelity text-to-speech model, with expressive multilingual voices and consistent speakers across long-form narration. Served through Azure Speech (preview).

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-voice-2

### microsoft/mai-voice-2-flash

Microsoft AI's low-latency text-to-speech model for voice agents and assistants, with the expressive multilingual voices of MAI-Voice-2 at twice the speed.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-voice-2-flash

### microsoft/mai-voice-2.1

MAI-Voice-2.1 is our highest-fidelity, most expressive text-to-speech model, delivering rich, natural speech across 23 languages. It extends the MAI-Voice family with broad multilingual coverage, gated instant voice cloning, and strong long-form generation capabilities. With its detailed prosody, nuanced expressiveness, and studio-grade audio quality, MAI-Voice-2.1 is ideal for experiences where maximum voice quality and fidelity are required - long-form narration, brand-defining audio etc.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-voice-2.1

### microsoft/mai-voice-2.1-flash

MAI-Voice-2.1-Flash is a text-to-speech model built for fast, low-latency generation. It produces high- fidelity, natural, and expressive speech across 23 languages and supports gated instant voice cloning, all while being optimized for real-time responsiveness. Its human-like intonation, rhythm, and emotional nuance make it ideal for voice agents, assistants, and other interactive scenarios where latency and cost are critical.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mai-voice-2.1-flash

### minimax/minimax-h3

H3 is a next-generation open-weights, general-purpose multimodal video model. Rather than being limited to specialized tasks such as generating, editing, or referencing, H3 understands multimodal contexts that bring together text, images, video, and audio. This enables it to interpret creative intent in a unified way and deliver more natural, coherent generation and expression.

- **Details:** https://vercel.com/ai-gateway/models/minimax-h3

### minimax/minimax-h3-max

MiniMax H3 Max is a next-generation general-purpose multimodal video model ranking #1 for overall quality, prompt understanding, and aesthetics in all evaluations, while generating a 5-second video in under 3 seconds.

- **Details:** https://vercel.com/ai-gateway/models/minimax-h3-max

### minimax/minimax-m2

MiniMax-M2 redefines efficiency for agents. It is a compact, fast, and cost-effective MoE model (230 billion total parameters with 10 billion active parameters) built for elite performance in coding and agentic tasks, all while maintaining powerful general intelligence.

- **Max output tokens:** 205K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2

### minimax/minimax-m2.1

MiniMax 2.1 is MiniMax's latest model, optimized specifically for robustness in coding, tool use, instruction following, and long-horizon planning.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.1

### minimax/minimax-m2.1-lightning

MiniMax-M2.1-lightning is a faster version of MiniMax-M2.1, offering the same performance but with significantly higher throughput (output speed ~100 TPS, MiniMax-M2 output speed ~60 TPS).

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.1-lightning

### minimax/minimax-m2.5

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. It is capable of handling the entire development process of various complex systems. It covers full-stack projects across multiple platforms including Web, Android, iOS, Windows, and Mac, encompassing server-side APIs, functional logic, and databases.

- **Max output tokens:** 131K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.5

### minimax/minimax-m2.5-highspeed

M2.5 highspeed: Same performance, faster and more agile (output speed approximately 100 tps)

- **Max output tokens:** 131K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.5-highspeed

### minimax/minimax-m2.7

M2.7 delivers outstanding performance in real-world software engineering, including end-to-end full project delivery, log analysis and bug troubleshooting, code security, machine learning, and more.

- **Max output tokens:** 196.6K
- **Cached input cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.7

### minimax/minimax-m2.7-highspeed

M2.7 Highspeed: Same performance, faster and more agile (output speed approximately 100 tps)

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/minimax-m2.7-highspeed

### minimax/minimax-m3

MiniMax-M3 is a frontier-class foundation model that unites the three capabilities defining today's frontier: a 1M-token context window, frontier coding and agentic performance, and native multimodality — the first open-weight model to deliver all three in a single system.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/minimax-m3

### minimax/minimax-m3.1-flash

MiniMax-M3.1-Flash is MiniMax's M-series language model for agent reasoning, tool use, coding, and long-context tasks, with a 1,000,000-token context window. Compared with MiniMax-M3, it adds always-on reasoning and adjustable reasoning effort, allowing developers to tune reasoning effort for different tasks.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/minimax-m3.1-flash

### mistral/codestral-embed

Code embedding model that can embed code databases and repositories to power coding assistants.

- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/codestral-embed

### mistral/ministral-14b

Ministral 3 14B is the largest model in the Ministral 3 family, offering state-of-the-art capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart. Optimized for local deployment, it delivers high performance across diverse hardware, including local setups.

- **Max output tokens:** 256K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/ministral-14b

### mistral/ministral-3b

A compact, efficient model for on-device tasks like smart assistants and local analytics, offering low-latency performance.

- **Max output tokens:** 4K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/ministral-3b

### mistral/ministral-8b

A more powerful model with faster, memory-efficient inference, ideal for complex workflows and demanding edge applications.

- **Max output tokens:** 4K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/ministral-8b

### mistral/codestral

Mistral's cutting-edge language model for coding released end of July 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation.

- **Max output tokens:** 4K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/codestral

### mistral/mistral-embed

General-purpose text embedding model for semantic search, similarity, clustering, and RAG workflows.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mistral-embed

### mistral/mistral-large-3

Mistral Large 3 2512 is Mistral’s most capable model to date. It has a sparse mixture-of-experts architecture with 41B active parameters (675B total).

- **Max output tokens:** 256K
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mistral-large-3

### mistral/mistral-large-4

Mistral Large 4 is Mistral AI's open-weight, natively multimodal model, available in public preview with 1.05 trillion total parameters and 49 billion active parameters. It leads aggregated benchmarks among open-weight models from the US and Europe, with strengths in cyber defense, manufacturing, finance, and visual grounding.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.14/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mistral-large-4

### mistral/mistral-medium-3.5

Mistral's frontier-class multimodal model optimized for agentic and coding use cases.

- **Max output tokens:** 256K
- **Cached input cost:** $0.15/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mistral-medium-3.5

### mistral/mistral-nemo

12B model trained jointly by Mistral AI and NVIDIA, it significantly outperforms existing models smaller or similar in size.

- **Max output tokens:** 131.1K
- **Details:** https://vercel.com/ai-gateway/models/mistral-nemo

### mistral/mistral-small

Mistral Small currently runs Mistral Small 4, a multimodal model combining instruction following, reasoning, and coding with a 262,144-token context window.

- **Max output tokens:** 4K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/mistral-small

### mistral/voxtral-mini-transcribe-3

Fast, cost-efficient multilingual transcription with speaker diarization and vocabulary biasing, designed for high-volume audio processing.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/voxtral-mini-transcribe-3

### mistral/voxtral-small-transcribe-3

Mistral’s flagship transcription model, optimized for multilingual accuracy, speaker diarization, and domain-specific vocabulary.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/voxtral-small-transcribe-3

### mistral/voxtral-mini-tts-3

Generates natural, expressive speech with preset or cloned voices and inline controls for emotion and pace.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/voxtral-mini-tts-3

### mixedbread/toast-1

Built for knowledge-intensive questions, multi-step retrieval, and evidence synthesis. Toast 1 can work with any backend exposed through function calling.

- **Max output tokens:** 4K
- **Cached input cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/toast-1

### moonshotai/kimi-k2

Kimi K2 is a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Trained with the Muon optimizer, Kimi K2 achieves exceptional performance across frontier knowledge, reasoning, and coding tasks while being meticulously optimized for agentic capabilities.

- **Max output tokens:** 131.1K
- **Details:** https://vercel.com/ai-gateway/models/kimi-k2

### moonshotai/kimi-k2.5

kimi-k2.5 is Kimi's most versatile model to date, featuring a native multimodal architecture that supports both visual and text input, thinking and non-thinking modes, and dialogue and agent tasks.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k2.5

### moonshotai/kimi-k2.6

Kimi K2.6 demonstrates particularly strong performance in long-horizon coding tasks and produces professional-grade design with code and vision.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.16/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k2.6

### moonshotai/kimi-k2.7-code

Kimi-K2.7-Code is a coding model from Moonshot AI. It has  improved coding & agent performance over K2.6, more reasoning efficiency with less overthinking, and improved instruction following for long-horizon coding.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.19/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k2.7-code

### moonshotai/kimi-k2.7-code-highspeed

Kimi K2.7 Code HighSpeed is the high-speed version of Kimi K2.7 Code, the same model as Kimi K2.7 Code, but with an output speed of approximately 180 Tokens/s and up to 260 Tokens/s in short context scenarios, delivering a more extreme coding experience.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.38/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k2.7-code-highspeed

### moonshotai/kimi-k3

Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.18/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k3

### moonshotai/kimi-k3-fast

Fast version of Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window.

- **Max output tokens:** 1M
- **Cached input cost:** $0.45/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k3-fast

### moonshotai/kimi-k3.1-pro

Kimi K3.1 Pro is a frontier model from Moonshot AI for complex reasoning, coding, and agentic workflows, with support for vision, tool use, structured outputs, and long-context tasks.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k3.1-pro

### moonshotai/kimi-k3.1-pro-highspeed

Kimi K3.1 Pro Highspeed is the high-speed variant of Moonshot AI’s frontier Kimi K3.1 Pro model for complex reasoning, coding, and agentic workflows, with vision, tool use, structured outputs, and long-context support.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/kimi-k3.1-pro-highspeed

### morph/morph-v3-fast

Morph offers a specialized AI model that applies code changes suggested by frontier models (like Claude or GPT-4o) to your existing code files FAST - 4500+ tokens/second. It acts as the final step in the AI coding workflow. Supports 16k input tokens and 16k output tokens.

- **Max output tokens:** 16.4K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/morph-v3-fast

### morph/morph-v3-large

Morph offers a specialized AI model that applies code changes suggested by frontier models (like Claude or GPT-4o) to your existing code files FAST - 2500+ tokens/second. It acts as the final step in the AI coding workflow. Supports 16k input tokens and 16k output tokens.

- **Max output tokens:** 16.4K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/morph-v3-large

### nvidia/nemotron-3-nano-30b-a3b

NVIDIA Nemotron 3 Nano is an open reasoning model optimized for fast, cost-efficient inference. Built with a hybrid MoE and Mamba architecture and trained on NVIDIA-curated synthetic reasoning data, it delivers strong multi-step reasoning with stable latency and predictable performance for agentic and production workloads.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3-nano-30b-a3b

### nvidia/nemotron-3-ultra-550b-a55b

A 550B parameter (55B active) open reasoning model from NVIDIA, built for long-running agent workflows. It uses a hybrid Mamba-Transformer MoE architecture and supports a 1M token context window.

- **Max output tokens:** 262K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3-ultra-550b-a55b

### nvidia/nemotron-3.5-lightning

NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 is a large language model (LLM) trained by NVIDIA.

The model employs a hybrid Mixture-of-Experts architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. The Lightning 3.5 model is released alongside a number of speculative decoding methods for faster text generation. The model has 3B active parameters and 30B parameters in total.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3.5-lightning

### nvidia/nemotron-3.5-lightning-free

Nemotron 3.5 Lightning is an open-weight reasoning and chat model for long-running agents, sub-agent work, chatbots, and RAG systems. The model has 30B parameters with 3B active per token, in a hybrid Mixture-of-Experts design that interleaves Mamba-2 and MoE layers with a few attention layers. It has a 1M token context window and supports tool calling and structured outputs. The provider may retain prompts and responses for model improvement.

- **Max output tokens:** 65K
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3.5-lightning-free

### nvidia/nemotron-3-super-120b-a12b

NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. It delivers up to 7x higher throughput, providing fast, cost-efficient inference for agentic tasks. Additionally, a long context window gives the model long-term memory, preventing AI agents from losing focus on long, multi-step tasks and ensuring high-accuracy results. Fully open with weights, datasets, and recipes, Super allows easy customization and secure deployment anywhere.

- **Max output tokens:** 32K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3-super-120b-a12b

### nvidia/nemotron-3.5-super-vl

NVIDIA Nemotron 3.5 Super VL is a vision-language model for reasoning, coding, tool use, and document understanding. Its hybrid Mamba-2, attention, and Mixture-of-Experts architecture has 120B total parameters with 12B active per token. Baseten serves text and image inputs with up to one million tokens of context for long conversations and agent workflows.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-3.5-super-vl

### nvidia/nemotron-nano-12b-v2-vl

The model is an auto-regressive vision language model that uses an optimized transformer architecture. The model enables multi-image reasoning and video understanding, along with strong document intelligence, visual Q&A and summarization capabilities.

- **Max output tokens:** 131.1K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-nano-12b-v2-vl

### nvidia/nemotron-nano-9b-v2

NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be controlled via a system prompt. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so.\

- **Max output tokens:** 131.1K
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/nemotron-nano-9b-v2

### openai/gpt-5.1-codex-max

GPT‑5.1-Codex-Max is purpose-built for agentic coding.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.1-codex-max

### openai/gpt-5.1-codex-mini

GPT-5.1 Codex mini is a smaller, faster, and cheaper version of GPT-5.1 Codex.

- **Max output tokens:** 128K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.1-codex-mini

### openai/gpt-5.1-thinking

An upgraded version of GPT-5 that adapts thinking time more precisely to the question to spend more time on complex questions and respond more quickly to simpler tasks.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.1-thinking

### openai/gpt-5.2

GPT-5.2 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.

- **Max output tokens:** 128K
- **Cached input cost:** $0.17/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.2

### openai/gpt-5.2-pro

Version of GPT-5.2 that produces smarter and more precise responses.

- **Max output tokens:** 128K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.2-pro

### openai/gpt-5.2-codex

GPT‑5.2-Codex is a version of GPT‑5.2⁠ further optimized for agentic coding in Codex, including improvements on long-horizon work through context compaction, stronger performance on large code changes like refactors and migrations, improved performance in Windows environments, and significantly stronger cybersecurity capabilities.

- **Max output tokens:** 128K
- **Cached input cost:** $0.17/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.2-codex

### openai/gpt-5.3-codex

GPT-5.3-Codex advances both the frontier coding performance of GPT‑5.2-Codex and the reasoning and professional knowledge capabilities of GPT‑5.2, together in one model, which is also 25% faster. This enables it to take on long-running tasks that involve research, tool use, and complex execution.

- **Max output tokens:** 128K
- **Cached input cost:** $0.17/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.3-codex

### openai/gpt-5.4

GPT-5.4 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.

- **Max output tokens:** 128K
- **Cached input cost:** $0.25/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.4

### openai/gpt-5.4-mini

GPT-5.4 Mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.

- **Max output tokens:** 128K
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.4-mini

### openai/gpt-5.4-nano

GPT-5.4 Nano is designed for tasks where speed and cost matter most like classification, data extraction, ranking, and sub-agents.

- **Max output tokens:** 128K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.4-nano

### openai/gpt-5.4-pro

GPT-5.4 Pro uses more compute to think harder and provide consistently better answers. It's designed to tackle tough problems.

- **Max output tokens:** 128K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.4-pro

### openai/gpt-5.5

GPT‑5.5 understands what you’re trying to do faster and can carry more of the work itself. It excels at writing and debugging code, researching online, analyzing data, creating documents and spreadsheets, operating software, and moving across tools until a task is finished. Instead of carefully managing every step, you can give GPT‑5.5 a messy, multi-part task and trust it to plan, use tools, check its work, navigate through ambiguity, and keep going.

- **Max output tokens:** 272K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.5

### openai/gpt-5.5-pro



- **Max output tokens:** 128K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.5-pro

### openai/gpt-5.6-luna

GPT-5.6 Luna is a fast, affordable GPT-5.6 model that brings strong capability at the lowest cost in the series.

- **Max output tokens:** 128K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.6-luna

### openai/gpt-5.6-sol

GPT-5.6 Sol is the flagship of OpenAI's GPT-5.6 series, its most capable model for long-horizon agentic work across coding, biology, and cybersecurity.

- **Max output tokens:** 128K
- **Cached input cost:** $0.40/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.6-sol

### openai/gpt-5.6-terra

GPT-5.6 Terra is a balanced GPT-5.6 model for everyday work, with performance comparable to the previous generation at half the cost.

- **Max output tokens:** 128K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.6-terra

### openai/gpt-image-1-mini

A cost-efficient version of GPT Image 1. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.

- **Cached input cost:** $0.20/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-image-1-mini

### openai/gpt-image-1.5

GPT Image 1.5 is OpenAI's latest image generation model, with better instruction following and adherence to prompts.

- **Cached input cost:** $1.25/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-image-1.5

### openai/gpt-image-2

GPT Image 2 is OpenAI's state-of-the-art image generation model for fast, high-quality image generation and editing. It supports flexible image sizes and high-fidelity image inputs.

- **Cached input cost:** $1.25/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-image-2

### openai/gpt-image-2.5-flare

GPT Image 2.5 Flare generates images from text and image inputs, optimized for fast, high-quality everyday image generation.

- **Cached input cost:** $1.25/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-image-2.5-flare

### openai/gpt-image-2.5-sunburst

GPT Image 2.5 Sunburst generates and edits images from text and image inputs, with a focus on precise editing.

- **Cached input cost:** $1.25/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-image-2.5-sunburst

### openai/gpt-live-transcribe

GPT Live Transcribe is a low-latency streaming speech-to-text model for live audio, with tunable latency, context, keyword hints, and multiple language hints. Priced by audio duration.

- **Details:** https://vercel.com/ai-gateway/models/gpt-live-transcribe

### openai/gpt-oss-120b

Extremely capable general-purpose LLM with strong, controllable reasoning capabilities

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.35/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-oss-120b

### openai/gpt-oss-20b

A compact, open-weight language model optimized for low-latency and resource-constrained environments, including local and edge deployments.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.02/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-oss-20b

### openai/gpt-oss-safeguard-120b

GPT OSS Safeguard 120B is OpenAI’s open-weight safety model for content moderation and guardrail enforcement. It helps teams evaluate content against custom safety policies and build scalable, adaptable protections for production AI applications.

- **Max output tokens:** 16K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-oss-safeguard-120b

### openai/gpt-oss-safeguard-20b

OpenAI's first open weight reasoning model specifically trained for safety classification tasks. Fine-tuned from GPT-OSS, this model helps classify text content based on customizable policies, enabling bring-your-own-policy Trust & Safety AI where your own taxonomy, definitions, and thresholds guide classification decisions.

- **Max output tokens:** 65.5K
- **Cached input cost:** $0.04/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-oss-safeguard-20b

### openai/gpt-transcribe

GPT Transcribe is a high-accuracy speech-to-text model for completed audio files, streamed file transcripts, and committed turns in Realtime sessions. It supports context, keyword hints, and multiple language hints. Priced by audio duration.

- **Details:** https://vercel.com/ai-gateway/models/gpt-transcribe

### openai/gpt-4.1

GPT 4.1 is OpenAI's flagship model for complex tasks. It is well suited for problem solving across domains.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-4.1

### openai/gpt-4.1-mini

GPT 4.1 mini provides a balance between intelligence, speed, and cost that makes it an attractive model for many use cases.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-4.1-mini

### openai/gpt-4.1-nano

GPT-4.1 nano is the fastest, most cost-effective GPT 4.1 model.

- **Max output tokens:** 32.8K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-4.1-nano

### openai/gpt-4o

GPT-4o from OpenAI has broad general knowledge and domain expertise allowing it to follow complex instructions in natural language and solve difficult problems accurately. It matches GPT-4 Turbo performance with a faster and cheaper API.

- **Max output tokens:** 16.4K
- **Cached input cost:** $1.25/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-4o

### openai/gpt-4o-mini

GPT-4o mini from OpenAI is their most advanced and cost-efficient small model. It is multi-modal (accepting text or image inputs and outputting text) and has higher intelligence than gpt-3.5-turbo but is just as fast.

- **Max output tokens:** 16.4K
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-4o-mini

### openai/gpt-4o-mini-transcribe

GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio. It offers improvements to word error rate and better language recognition and accuracy compared to original Whisper models. Use it for more accurate transcripts.

- **Details:** https://vercel.com/ai-gateway/models/gpt-4o-mini-transcribe

### openai/gpt-4o-transcribe

GPT-4o Transcribe is a speech-to-text model that uses GPT-4o to transcribe audio. It offers improvements to word error rate and better language recognition and accuracy compared to original Whisper models. Use it for more accurate transcripts.

- **Details:** https://vercel.com/ai-gateway/models/gpt-4o-transcribe

### openai/gpt-5

GPT-5 is OpenAI's flagship language model that excels at complex reasoning, broad real-world knowledge, code-intensive, and multi-step agentic tasks.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5

### openai/gpt-5-mini

GPT-5 mini is a cost optimized model that excels at reasoning/chat tasks. It offers an optimal balance between speed, cost, and capability.

- **Max output tokens:** 128K
- **Cached input cost:** $0.03/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5-mini

### openai/gpt-5-nano

GPT-5 nano is a high throughput model that excels at simple instruction or classification tasks.

- **Max output tokens:** 128K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5-nano

### openai/gpt-5-pro

GPT-5 pro uses more compute to think harder and provide consistently better answers. Since GPT-5 pro is designed to tackle tough problems, some requests may take several minutes to finish.

- **Max output tokens:** 272K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5-pro

### openai/gpt-5-codex

GPT-5-Codex is a version of GPT-5 optimized for agentic coding tasks in Codex or similar environments.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5-codex

### openai/gpt-5.1-codex

GPT-5.1-Codex is a version of GPT-5.1 optimized for agentic coding tasks in Codex or similar environments.

- **Max output tokens:** 128K
- **Cached input cost:** $0.13/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-5.1-codex

### openai/gpt-6-astra

GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.

- **Max output tokens:** 128K
- **Cached input cost:** $1.00/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-6-astra

### openai/gpt-6-luna

GPT-6 Luna is OpenAI's efficient reasoning model for focused, high-volume tasks. It accepts text and images, generates text, and supports a 1,050,000-token context window with up to 128,000 output tokens. Configurable reasoning effort, structured outputs, prompt caching, and Responses API tools support cost-sensitive applications, coding tasks, and automated workflows.

- **Max output tokens:** 128K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-6-luna

### openai/gpt-6-luna-decisions

GPT-6 Luna Decisions evaluates classification and scoring questions against shared text and image input using OpenAI's Decisions API. It supports predicate, choice, and score questions with probabilities and confidence values.

- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-6-luna-decisions

### openai/gpt-6-sol

GPT-6 Sol is OpenAI's reasoning model for complex coding and agentic workflows. It accepts text and images, generates text, and supports a 1,050,000-token context window with up to 128,000 output tokens. Configurable reasoning effort, structured outputs, prompt caching, and Responses API tools make it suited to software development and multi-step automation.

- **Max output tokens:** 128K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-6-sol

### openai/gpt-6.1-sol

GPT-6.1 Sol is OpenAI's reasoning model for complex coding, computer use, and professional work. It accepts text and images, generates text, and supports a 1,050,000-token context window with up to 128,000 output tokens. Configurable reasoning effort, structured outputs, prompt caching, and Responses API tools support multi-step workflows.

- **Max output tokens:** 128K
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-6.1-sol

### openai/gpt-6.1-terra

GPT-6.1 Terra is OpenAI’s model for visual reasoning, precise computer use, and natural writing. It accepts text and images and generates text, with configurable reasoning effort, structured outputs, tool calling through the Responses API, and implicit and explicit prompt caching.

- **Max output tokens:** 128K
- **Cached input cost:** $0.05/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/gpt-6.1-terra

### openai/gpt-live-1

Our premier model for natural, expressive voice conversations with smooth interruption handling.

- **Details:** https://vercel.com/ai-gateway/models/gpt-live-1

### openai/gpt-realtime-mini

GPT-Realtime mini is capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.

- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-realtime-mini

### openai/gpt-realtime-1.5

GPT-Realtime-1.5 is our flagship audio model for voice agents and customer support.

- **Cached input cost:** $0.40/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-realtime-1.5

### openai/gpt-realtime-2

GPT Realtime 2 is our most capable realtime voice model. It supports speech-to-speech interactions with configurable reasoning effort, stronger instruction following, and more reliable tool use for complex voice-agent workflows.

- **Cached input cost:** $0.40/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-realtime-2

### openai/gpt-realtime-2.1

GPT-Realtime-2.1 updates GPT-Realtime-2 with improved alphanumeric recognition, silence and noise handling, and interruption behavior. It supports speech-to-speech interactions with configurable reasoning effort, instruction following, and tool use for complex voice-agent workflows.

- **Max output tokens:** 32K
- **Cached input cost:** $0.40/1M
- **Details:** https://vercel.com/ai-gateway/models/gpt-realtime-2.1

### openai/gpt-realtime-whisper

GPT Realtime Whisper is a streaming speech-to-text model for applications that need low-latency transcript deltas from live audio. It is designed for realtime use cases where developers need to tune latency and accuracy. GPT Realtime Whisper is priced by audio duration rather than text tokens.

- **Details:** https://vercel.com/ai-gateway/models/gpt-realtime-whisper

### openai/o1

o1 is OpenAI's flagship reasoning model, designed for complex problems that require deep thinking. It provides strong reasoning capabilities with improved accuracy for complex multi-step tasks.

- **Max output tokens:** 100K
- **Cached input cost:** $7.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/o1

### openai/o3

OpenAI's o3 is their most powerful reasoning model, setting new state-of-the-art benchmarks in coding, math, science, and visual perception. It excels at complex queries requiring multi-faceted analysis, with particular strength in analyzing images, charts, and graphics.

- **Max output tokens:** 100K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/o3

### openai/o3-pro

The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers.

- **Max output tokens:** 100K
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/o3-pro

### openai/o3-mini

o3-mini is OpenAI's most recent small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini.

- **Max output tokens:** 100K
- **Cached input cost:** $0.55/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/o3-mini

### openai/o4-mini

OpenAI's o4-mini delivers fast, cost-efficient reasoning with exceptional performance for its size, particularly excelling in math (best-performing on AIME benchmarks), coding, and visual tasks.

- **Max output tokens:** 100K
- **Cached input cost:** $0.28/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/o4-mini

### openai/text-embedding-3-large

OpenAI's most capable embedding model for both english and non-english tasks.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/text-embedding-3-large

### openai/text-embedding-3-small

OpenAI's improved, more performant version of their ada embedding model.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/text-embedding-3-small

### openai/text-embedding-ada-002

OpenAI's legacy text embedding model.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/text-embedding-ada-002

### openai/tts-1

TTS is a model that converts text to natural sounding spoken text.

- **Details:** https://vercel.com/ai-gateway/models/tts-1

### openai/tts-1-hd

TTS is a model that converts text to natural sounding spoken text. The tts-1-hd model is optimized for high quality text-to-speech use cases.

- **Details:** https://vercel.com/ai-gateway/models/tts-1-hd

### openai/whisper-1

Whisper is a general-purpose speech recognition model, trained on a large dataset of diverse audio. You can also use it as a multitask model to perform multilingual speech recognition as well as speech translation and language identification.

- **Details:** https://vercel.com/ai-gateway/models/whisper-1

### particle/particle-1

Particle-1 is Particle AI’s model for agentic coding and software engineering, post-trained on DeepSeek V4.1 Flash to complete tasks with fewer tokens. It supports text and image inputs, reasoning, tool use, and implicit prompt caching.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/particle-1

### perplexity/pplx-decider-flash-v1.2

Perplexity Decider Flash V1.2 is a 4B multimodal decision model for high-volume classification, routing, labeling, and filtering. It evaluates text, structured JSON, and images against yes/no questions, custom choices, and ordered rubrics, returning typed answers with probabilities. It supports a 256K context window.

- **Details:** https://vercel.com/ai-gateway/models/pplx-decider-flash-v1.2

### perplexity/pplx-decider-v1.2

Perplexity Decider V1.2 is a 27B multimodal decision model for classification, routing, and scoring. It evaluates text, structured JSON, and images against yes/no questions, custom choices, and ordered rubrics, returning typed answers with probabilities. It supports multiple questions over shared content and a 256K context window.

- **Details:** https://vercel.com/ai-gateway/models/pplx-decider-v1.2

### perplexity/pplx-embed-v1-0.6b

pplx-embed-v1 is a state-of-the-art text embedding model optimized for real-world, web-scale retrieval tasks.

- **Details:** https://vercel.com/ai-gateway/models/pplx-embed-v1-0.6b

### perplexity/pplx-embed-v1-4b

pplx-embed-v1 is a state-of-the-art text embedding model optimized for real-world, web-scale retrieval tasks.

- **Details:** https://vercel.com/ai-gateway/models/pplx-embed-v1-4b

### perplexity/sonar

Perplexity's lightweight offering with search grounding, quicker and cheaper than Sonar Pro.

- **Max output tokens:** 8K
- **Details:** https://vercel.com/ai-gateway/models/sonar

### poolside/laguna-s-2.1

Laguna S 2.1 is Poolside's new open-weight model for agentic coding and long-horizon work.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/laguna-s-2.1

### poolside/laguna-s-2.1-free

Laguna S 2.1 is Poolside's new open-weight model for agentic coding and long-horizon work.

- **Max output tokens:** 32.8K
- **Details:** https://vercel.com/ai-gateway/models/laguna-s-2.1-free

### prodia/flux-fast-schnell

Lightning-fast image generation

- **Details:** https://vercel.com/ai-gateway/models/flux-fast-schnell

### quiverai/arrow-2

SVG generation model with agentic refinement and tool calling.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.40/1M
- **Details:** https://vercel.com/ai-gateway/models/arrow-2

### quiverai/arrow-2-telos

Higher-refinement Arrow 2 variant combining Arrow's speed with frontier-model reasoning.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.60/1M
- **Details:** https://vercel.com/ai-gateway/models/arrow-2-telos

### recraft/recraft-v2

Recraft V2 is an image generation model released in March 2024 and the first model trained from scratch by Recraft. With 20 billion parameters, it was a breakthrough in human anatomical accuracy and the first to support brand consistency and brand color inputs. It also introduced vector image generation (SVG output), as well as minimalistic icon and illustration styles.

- **Image cost:** $0.02/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v2

### recraft/recraft-v3

V3 introduced major advances in photorealism and text rendering. It was the first Recraft model to generate mid-size text accurately and, as of 2025, is the only model capable of placing text at specific positions in an image.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v3

### recraft/recraft-v4

The model delivers strong photorealism, including realistic skin rendering and natural textures, while avoiding common synthetic artifacts. It produces more distinctive lighting, composition, diverse subjects, contemporary styling, and carefully considered scene elements. For illustration, it generates original characters and forms with sophisticated and unexpected color combinations.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4

### recraft/recraft-v4-pro

The model delivers strong photorealism, including realistic skin rendering and natural textures, while avoiding common synthetic artifacts. It produces more distinctive lighting, composition, diverse subjects, contemporary styling, and carefully considered scene elements. For illustration, it generates original characters and forms with sophisticated and unexpected color combinations.

- **Image cost:** $0.25/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4-pro

### recraft/recraft-v4.1

V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4.1

### recraft/recraft-v4.1-flash

Recraft V4.1 Flash is a fast, low-cost model with quality close to Recraft V4.1, built for high-volume generation and rapid iteration. Strong use cases include photography, portraits, atmosphere, complex scenes, mixed media, and logo concepts.

- **Image cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4.1-flash

### recraft/recraft-v4.1-pro

V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Pro generates higher-resolution images for when the idea deserves more room.

- **Image cost:** $0.21/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4.1-pro

### recraft/recraft-v4.1-utility

V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Utility is designed for when restraint is the aesthetic choice, with flat lighting, front-facing composition, and simple, controlled scenes.

- **Image cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4.1-utility

### recraft/recraft-v4.1-utility-pro

V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Utility is designed for when restraint is the aesthetic choice, with flat lighting, front-facing composition, and simple, controlled scenes.

- **Image cost:** $0.21/1M
- **Details:** https://vercel.com/ai-gateway/models/recraft-v4.1-utility-pro

### reflection/beam

Beam is Reflection’s reasoning model for coding and agentic workflows, combining strong coding capabilities with fast, low-latency inference.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.04/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/beam

### sakana/fugu-max

Fugu Max orchestrates our largest pool of models to push the cost–performance Pareto frontier. It dynamically finds efficient combinations of expert agents, aiming to improve model performance and cost efficiency together.

- **Max output tokens:** 1M
- **Cached input cost:** $0.25/1M
- **Details:** https://vercel.com/ai-gateway/models/fugu-max

### sakana/fugu-ultra

Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Fugu Ultra focuses on maximizing performance, for a higher cost. It routes between one to three agents, depending on the problem.

- **Max output tokens:** 1M
- **Cached input cost:** $0.50/1M
- **Details:** https://vercel.com/ai-gateway/models/fugu-ultra

### sakana/fugu-ultra-v2

Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Fugu Ultra focuses on maximizing performance, for a higher cost.

- **Max output tokens:** 1M
- **Cached input cost:** $0.50/1M
- **Details:** https://vercel.com/ai-gateway/models/fugu-ultra-v2

### sakana/namazu

Sakana Namazu is a Japanese-specialized LLM that combines a deep understanding of Japanese culture and business customs with high-performance language capabilities. Built on the open model Kimi K2.6 and refined with Sakana AI's in-house data for Japanese language and business workflows, it handles complex tasks using web search and code execution. Unlike Fugu, which orchestrates multiple frontier models, Sakana Namazu provides a single in-house model as an API.

- **Max output tokens:** 256K
- **Cached input cost:** $0.15/1M
- **Details:** https://vercel.com/ai-gateway/models/namazu

### scaledown/scaledown-classify-1

ScaleDown Classify-1 is a task-specific small language model for text classification. It assigns text to custom labels and returns probability scores, supporting intent detection, content categorization, and routing workflows. Its deterministic, encoder-only approach produces consistent classification results without sampling variability.

- **Details:** https://vercel.com/ai-gateway/models/scaledown-classify-1

### scaledown/scaledown-extract-1

ScaleDown Extract-1 is a task-specific language model for extracting structured data from unstructured text using a custom JSON schema. It identifies domain-specific entities with confidence scores, character offsets, and surrounding context, and supports nested objects and arrays for workflows such as invoice processing, contract analysis, and record parsing.

- **Max output tokens:** 1M
- **Details:** https://vercel.com/ai-gateway/models/scaledown-extract-1

### scaledown/scaledown-summarize-1

ScaleDown Summarize-1 is a task-specific language model for abstractive summarization of long documents, transcripts, and conversations. It rewrites source text into concise summaries and accepts instructions for format, language, length, and focus, making it useful for document processing, meeting summaries, and support workflows.

- **Max output tokens:** 1K
- **Details:** https://vercel.com/ai-gateway/models/scaledown-summarize-1

### stealth/glyph-cluster

Glyph Cluster is an anonymous reasoning model in early access. Prompts and outputs will be retained by the provider.

- **Max output tokens:** 256K
- **Details:** https://vercel.com/ai-gateway/models/glyph-cluster

### stepfun/step-3.7-flash

StepFun’s flagship multimodal reasoning model. Powered by a 198B-parameter / 11B-activation sparse MoE architecture, with native support for image and video understanding.

- **Max output tokens:** 256K
- **Cached input cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/step-3.7-flash

### stepfun/step-5-preview

Step 5 Preview is StepFun’s flagship AI model for agentic coding, professional knowledge work, and financial analysis. Built on a sparse Mixture-of-Experts architecture with 600B total parameters and 27B active per token, it combines a 1M-token context window with vision input. From building interactive applications and debugging code to conducting research and producing analytical reports, it supports complex workflows that require sustained reasoning, tool use, and iterative refinement.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.05/1M
- **Details:** https://vercel.com/ai-gateway/models/step-5-preview

### tencent/hy3



- **Max output tokens:** 262.1K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/hy3

### tencent/hy-mt2-lite

Hy-MT2-Lite is Tencent Cloud Hunyuan’s lightweight 1.8B translation model, combining efficient performance with an 8K-token context window and broad multilingual coverage. It supports translation across 33 languages and five ethnic Chinese and dialect variants, delivering strong results on benchmarks including FLORES-200 and WMT25. Designed for professional and real-world business use, it reliably follows instructions for structured, delimiter-preserving, context-aware, glossary-guided, and style-specific translation.

- **Max output tokens:** 4K
- **Details:** https://vercel.com/ai-gateway/models/hy-mt2-lite

### tencent/hy-mt2-plus

Hy-MT2-Plus is Tencent Cloud Hunyuan’s 7B translation model, offering an 8K-token context window and broad multilingual coverage. Optimized for translation across 33 languages and five ethnic Chinese and dialect variants, it delivers industry-leading performance on benchmarks including FLORES-200 and WMT25. It excels in professional domains and real-world business scenarios, with strong instruction-following for structured, delimiter-preserving, context-aware, glossary-guided, and style-specific translation.

- **Max output tokens:** 4K
- **Details:** https://vercel.com/ai-gateway/models/hy-mt2-plus

### tencent/hy-mt2-pro

Hy-MT2-Pro is Tencent Cloud Hunyuan’s flagship translation model, featuring 30B-A3B parameters and an 8K-token context window. It delivers bidirectional translation across 33 core language pairs, plus five minority-language and dialect pairs. With industry-leading performance on FLORES-200 and WMT25, it excels in specialized domains and real-world business scenarios while supporting structured, contextual, terminology-aware, style-specific, and delimiter-preserving translation.

- **Max output tokens:** 4K
- **Details:** https://vercel.com/ai-gateway/models/hy-mt2-pro

### tencent/hy4-preview

Tencent Hy4 preview is Tencent Hy’s open-source large language model, featuring 770B total parameters, 49B active parameters, and a context window exceeding 1 million tokens. Built for real-world productivity, it supports long-horizon coding, cross-document analysis, office content creation, game development, and scientific reasoning, making it well-suited to complex agents and multi-step professional workflows.

- **Max output tokens:** 64K
- **Cached input cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/hy4-preview

### thinkingmachines/inkling

Inkling is a multimodal MoE model (975B total, 41B active, 256k context) reasoning over text, image, and audio inputs.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.16/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/inkling

### thinkingmachines/inkling-small

Inkling-Small is a lighter-weight model with 12B active parameters, trained with a similar recipe, to Inkling that achieves strong performance with even lower cost and latency.

- **Max output tokens:** 1M
- **Cached input cost:** $0.06/1M
- **Details:** https://vercel.com/ai-gateway/models/inkling-small

### topaz/proteus

Proteus is Topaz Labs’ video enhancement model for detail-preserving upscaling, denoising, and restoration. It enhances an existing video with automatic adjustments or fine control over detail recovery, compression artifacts, noise, blur, and grain. An input video and output resolution are required, supporting workflows that improve footage quality while preserving its source characteristics.

- **Details:** https://vercel.com/ai-gateway/models/proteus

### topaz/starlight-precise-2.6

Starlight Precise 2.6 is Topaz Labs’ generative video enhancement model for upscaling and restoration with high fidelity. It improves faces, textures, and fine details while maintaining temporal consistency across frames. An input video and accurate source metadata are required, making it suited to refining generated footage and restoring archival video for higher-resolution delivery.

- **Details:** https://vercel.com/ai-gateway/models/starlight-precise-2.6

### topaz/wonder-3.5

Wonder 3.5 is Topaz Labs’ image enhancement model for generative upscaling and detail recovery. It repairs and enhances an existing JPEG, PNG, or TIFF image, with controls for enhancement strength, output dimensions, and grain. An input image is required, making it suited to restoring low-quality photos and refining images for larger displays or print.

- **Details:** https://vercel.com/ai-gateway/models/wonder-3.5

### typesafe-ai/jev

Jev is TypeSafe AI’s System One decision model for fast, structured decisions in software. It evaluates shared state against typed questions and returns choices, scores, and boolean probabilities, supporting classification, routing, rubric-based assessment, and automated verification. Multiple questions can be evaluated in parallel within a single request.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/jev

### typesafe-ai/jev-1.13.0

Jev 1.13.0 is TypeSafe AI’s System One evaluation model for fast, structured decisions in software. It evaluates shared state against typed questions and returns choices, scores, and boolean probabilities, supporting classification, routing, rubric-based assessment, and automated verification. Multiple questions can be evaluated in parallel within a single request.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/jev-1.13.0

### typesafe-ai/jev-1.14.0

Jev 1.14.0 is TypeSafe AI’s System One evaluation model for fast, structured decisions in software. It evaluates shared state against typed questions and returns choices, scores, and boolean probabilities, supporting classification, routing, rubric-based assessment, and automated verification. Multiple questions can be evaluated in parallel within a single request.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/jev-1.14.0

### voyage/voyage-3.5

Voyage AI's embedding model optimized for general-purpose and multilingual retrieval quality.

- **Details:** https://vercel.com/ai-gateway/models/voyage-3.5

### voyage/voyage-3.5-lite

Voyage AI's embedding model optimized for latency and cost.

- **Details:** https://vercel.com/ai-gateway/models/voyage-3.5-lite

### voyage/voyage-4

Optimized for general-purpose and multilingual retrieval quality. All embeddings created with the 4 series are compatible with each other.

- **Details:** https://vercel.com/ai-gateway/models/voyage-4

### voyage/voyage-4-large

The best general-purpose and multilingual retrieval quality. All embeddings created with the 4 series are compatible with each other.

- **Details:** https://vercel.com/ai-gateway/models/voyage-4-large

### voyage/voyage-4-lite

Optimized for latency and cost. All embeddings created with the 4 series are compatible with each other.

- **Details:** https://vercel.com/ai-gateway/models/voyage-4-lite

### voyage/voyage-code-2

Voyage AI's embedding model optimized for code retrieval (17% better than alternatives). This is the previous generation of code embeddings models.

- **Details:** https://vercel.com/ai-gateway/models/voyage-code-2

### voyage/voyage-code-3

Voyage AI's embedding model optimized for code retrieval.

- **Details:** https://vercel.com/ai-gateway/models/voyage-code-3

### voyage/voyage-finance-2

Voyage AI's embedding model optimized for finance retrieval and RAG.

- **Details:** https://vercel.com/ai-gateway/models/voyage-finance-2

### voyage/voyage-law-2

Voyage AI's embedding model optimized for legal retrieval and RAG.

- **Details:** https://vercel.com/ai-gateway/models/voyage-law-2

### voyage/rerank-2.5

A generalist reranker optimized for quality with instruction-following and multilingual support.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-2.5

### voyage/rerank-2.5-lite

A generalist reranker optimized for both latency and quality with instruction-following and multilingual support.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-2.5-lite

### voyage/rerank-3

A general-purpose reranker optimized for retrieval accuracy.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-3

### voyage/rerank-3-lite

A fast, cost-effective reranker optimized for latency-sensitive retrieval.

- **Max output tokens:** 32K
- **Details:** https://vercel.com/ai-gateway/models/rerank-3-lite

### voyage/voyage-3-large

Voyage AI's embedding model with the best general-purpose and multilingual retrieval quality.

- **Details:** https://vercel.com/ai-gateway/models/voyage-3-large

### spacexai/grok-4.20-non-reasoning-beta

Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-non-reasoning-beta

### spacexai/grok-4.20-reasoning-beta

Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-reasoning-beta

### spacexai/grok-4.20-multi-agent-beta

Multiple agents collaborate in parallel to perform deep research tasks.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-multi-agent-beta

### spacexai/grok-4.20-multi-agent

Multiple agents collaborate in parallel to perform deep research tasks.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-multi-agent

### spacexai/grok-4.20-non-reasoning

Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherence, delivering consistently precise and truthful responses.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-non-reasoning

### spacexai/grok-4.20-reasoning

Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherence, delivering consistently precise and truthful responses.

- **Max output tokens:** 2M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.20-reasoning

### spacexai/grok-4.3

Grok 4.3 is a new model matching the scale of Grok 4.20 with an improved architecture and a December 2025 knowledge cutoff.

- **Max output tokens:** 1M
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.3

### spacexai/grok-4.5

SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.

- **Max output tokens:** 500K
- **Cached input cost:** $0.30/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.5

### spacexai/grok-4.6

Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application or work artifact.

- **Max output tokens:** 500K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.6

### spacexai/grok-4.7

Grok 4.7 is SpaceXAI’s advanced AI model for coding and professional knowledge work, built to tackle complex, multi-hour tasks with improved self-verification and long-context handling. It strengthens software engineering, document creation, and presentation workflows while maintaining Grok 4.6’s speed and pricing.

- **Max output tokens:** 500K
- **Cached input cost:** $0.50/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-4.7

### spacexai/grok-build-0.1

xAI's fast coding model trained specifically for agentic coding.

- **Max output tokens:** 256K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-build-0.1

### spacexai/grok-imagine-video

State-of-the-art video generation across quality, cost, and latency. Grok Imagine is x.AI's most powerful video-audio generative model yet. Bring an image to life, start from a simple text prompt, or even refine a complex cinematic sequence.

- **Details:** https://vercel.com/ai-gateway/models/grok-imagine-video

### spacexai/grok-imagine-image

Generate high-quality images from text prompts with xAI's imagine API.

- **Image cost:** $0.02/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-imagine-image

### spacexai/grok-imagine-image-2.0



- **Image cost:** $0.06/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-imagine-image-2.0

### spacexai/grok-imagine-video-1.5



- **Details:** https://vercel.com/ai-gateway/models/grok-imagine-video-1.5

### spacexai/grok-imagine-video-1.5-lite

Grok Imagine Video 1.5 Lite is xAI's lightweight video generation model for turning text prompts and images into clips with native audio. It supports videos from 1 to 15 seconds in 480p, 720p, and 1080p, with seven aspect ratios for formats including widescreen 16:9 and vertical 9:16.

- **Details:** https://vercel.com/ai-gateway/models/grok-imagine-video-1.5-lite

### spacexai/grok-stt

Transcribe audio to text in 25 languages with batch and streaming modes.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-stt

### spacexai/grok-tts

Generate speech with 5 expressive voices, speech tags, and telephony codecs.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-tts

### spacexai/grok-voice-3.0

Grok Voice 3.0 is xAI’s realtime speech-to-speech model for interactive voice agents. It streams audio and text bidirectionally, supports multilingual conversations and function calling, and includes reasoning enabled by default for complex instructions and nuanced dialogue.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-voice-3.0

### spacexai/grok-voice-think-fast-2.0

Grok Voice Think Fast 2.0, our next-generation voice model with improved intelligence, transcription accuracy, and conversational capabilities.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/grok-voice-think-fast-2.0

### xiaomi/mimo-v2.5

A native full-modal model supporting text, image, video, and audio understanding, with powerful Agent capabilities.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/mimo-v2.5

### xiaomi/mimo-v2.5-pro

MiMo V2.5 Pro delivers significant improvements over its predecessor, MiMo-V2-Pro, in general agentic capabilities, complex software engineering, and long-horizon tasks. MiMo-V2.5-Pro is a 1.02T-parameter Mixture-of-Experts model with 42B active parameters, built on a hybrid-attention architecture with a 1M-token context window.

- **Max output tokens:** 131K
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/mimo-v2.5-pro

### xiaomi/mimo-v2.6-flash

MiMo V2.6 Flash is Xiaomi's efficient multimodal reasoning model for coding, automation, and everyday agent workflows. It accepts text, images, audio, and video within a 1M-token context window, with up to 128K tokens of output. Deep thinking, tool calling, JSON mode, and prompt caching make it suitable for applications that need multimodal understanding at a lower token cost.

- **Max output tokens:** 1.1M
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/mimo-v2.6-flash

### xiaomi/mimo-v2.6-pro

MiMo V2.6 Pro is Xiaomi's flagship multimodal reasoning model for complex software engineering, long-running agent tasks, and professional workflows. It supports text, image, audio, and video inputs with a 1M-token context window and up to 128K tokens of output. Deep thinking, tool calling, JSON mode, and prompt caching support applications that combine large inputs with multi-step reasoning.

- **Max output tokens:** 1.1M
- **Cached input cost:** $0.00/1M
- **Details:** https://vercel.com/ai-gateway/models/mimo-v2.6-pro

### xiaomi/mimo-v2.6-pro-ultraspeed

MiMo V2.6 Pro UltraSpeed is Xiaomi's accelerated inference offering for MiMo V2.6 Pro, designed for interactive agents and workflows where response latency matters. It combines multimodal understanding of text, images, audio, and video with a 1M-token context window and up to 128K tokens of output. It supports deep thinking, tool calling, JSON mode, and prompt caching.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.04/1M
- **Details:** https://vercel.com/ai-gateway/models/mimo-v2.6-pro-ultraspeed

### zai/glm-4.5

GLM-4.5 and GLM-4.5-Air are our latest flagship models, purpose-built as foundational models for agent-oriented applications. Both leverage a Mixture-of-Experts (MoE) architecture. GLM-4.5 has a total parameter count of 355B with 32B active parameters per forward pass, while GLM-4.5-Air adopts a more streamlined design with 106B total parameters and 12B active parameters.

- **Max output tokens:** 96K
- **Cached input cost:** $0.11/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-4.5

### zai/glm-4.5-air

GLM-4.5 and GLM-4.5-Air are our latest flagship models, purpose-built as foundational models for agent-oriented applications. Both leverage a Mixture-of-Experts (MoE) architecture. GLM-4.5 has a total parameter count of 355B with 32B active parameters per forward pass, while GLM-4.5-Air adopts a more streamlined design with 106B total parameters and 12B active parameters.

- **Max output tokens:** 96K
- **Cached input cost:** $0.03/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-4.5-air

### zai/glm-4.5v

Built on the GLM-4.5-Air base model, GLM-4.5V inherits proven techniques from GLM-4.1V-Thinking while achieving effective scaling through a powerful 106B-parameter MoE architecture.

- **Max output tokens:** 16.4K
- **Cached input cost:** $0.11/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-4.5v

### zai/glm-4.6

As the latest iteration in the GLM series, GLM-4.6 achieves comprehensive enhancements across multiple domains, including real-world coding, long-context processing, reasoning, searching, writing, and agentic applications.

- **Max output tokens:** 202.8K
- **Cached input cost:** $0.11/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-4.6

### zai/glm-4.7

GLM-4.7 is Z.ai’s latest flagship model, with major upgrades focused on two key areas: stronger coding capabilities and more stable multi-step reasoning and execution.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.08/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-4.7

### zai/glm-4.7-flash

GLM-4.7-Flash balances high performance with efficiency, making it the perfect lightweight deployment option. Beyond coding, it is also recommended for creative writing, translation, long-context tasks, and roleplay.

- **Max output tokens:** 131K
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-4.7-flash

### zai/glm-4.7-flashx

 GLM-4.7-Flash balances high performance with efficiency, making it the perfect lightweight deployment option. 

- **Max output tokens:** 128K
- **Cached input cost:** $0.01/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-4.7-flashx

### zai/glm-5

GLM-5 is Zai’s new-generation flagship foundation model, designed for Agentic Engineering, capable of providing reliable productivity in complex system engineering and long-range Agent tasks. In terms of Coding and Agent capabilities, GLM-5 has achieved state-of-the-art (SOTA) performance in open source, with its usability in real programming scenarios approaching that of Claude Opus 4.5.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.20/1M
- **Zero data retention:** available
- **HIPAA compliant:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5

### zai/glm-5-turbo

GLM 5 Turbo is a foundation model deeply optimized for the OpenClaw scenario. It has been specifically optimized for the core requirements of OpenClaw tasks since the training phase, enhancing key capabilities such as tool invocation, command following, timed and persistent tasks, and long-chain execution.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.24/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-5-turbo

### zai/glm-5.1

GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks. Unlike previous models built around minute-level interactions, GLM-5.1 can work independently and continuously on a single task for more than 8 hours—autonomously planning, executing, and improving itself throughout the process—ultimately delivering complete, engineering-grade results.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.26/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-5.1

### zai/glm-5.2

GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.10/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5.2

### zai/glm-5.2-fast

Fast version of GLM 5.2 with 120-250 TPS.

- **Max output tokens:** 128K
- **Cached input cost:** $0.21/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5.2-fast

### zai/glm-5.3

GLM 5.3 delivers comprehensive advancements in complex software engineering and agent capabilities. It uses the same base model as GLM-5.2, with all improvements driven by post-training.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.07/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5.3

### zai/glm-5.3-fast

Speed-optimized version of Z.AI’s GLM-5.3 agentic coding model, built for responsive, real-time workloads. It supports advanced coding, long-horizon task execution, testing, vulnerability discovery, and cybersecurity analysis, making it ideal for developer tools, coding agents, and complex multi-step engineering workflows.

- **Max output tokens:** 262.1K
- **Cached input cost:** $0.21/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5.3-fast

### zai/glm-5.3-flash

GLM-5.3-Flash is Z.ai’s native multimodal coding model, featuring 320B total parameters, 18B activated parameters, and a 1M-token context window. Its efficient hybrid attention architecture supports visual coding, tool use, and end-to-end professional workflows across code, browsers, documents, and graphical interfaces.

- **Max output tokens:** 1.0M
- **Cached input cost:** $0.01/1M
- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/glm-5.3-flash

### zai/glm-5.3-flashx

GLM-5.3-FlashX is the high-speed serving option for Z.ai’s native multimodal coding model, featuring 320B total parameters, 18B activated parameters, and a 1M-token context window. Its efficient hybrid attention architecture supports visual coding, tool use, and end-to-end professional workflows across code, browsers, documents, and graphical interfaces.

- **Max output tokens:** 131.1K
- **Cached input cost:** $0.07/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-5.3-flashx

### zai/glm-5v-turbo

GLM-5V-Turbo is Z.AI’s first multimodal coding foundation model, built for vision-based coding tasks. It can natively process multimodal inputs such as images, video, and text, while also excelling at long-horizon planning, complex coding, and action execution. Deeply optimized for agent workflows, it works seamlessly with agents such as Claude Code and OpenClaw to complete the full loop of “understand the environment → plan actions → execute tasks”.

- **Max output tokens:** 128K
- **Cached input cost:** $0.24/1M
- **Details:** https://vercel.com/ai-gateway/models/glm-5v-turbo

### exa/search

Search the web with Exa's neural search API, returning LLM-optimized, ranked results with content excerpts. Use it as a built-in tool through the AI Gateway to give any model access to real-time web information.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/search

### perplexity/search

Search the web using Perplexity's Search API for real-time information, news, research papers, and articles. Provides ranked search results with advanced filtering options including domain, language, date range, and recency filters.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/search

### parallel/search

Search the web with Parallel's Search API for LLM-optimized excerpts. Every mode costs the same ($5 per 1,000 searches, plus $1 per 1,000 results beyond 10): turbo and fast trade excerpt depth for lower latency (turbo supports English and Japanese only), basic is the balanced default, and advanced is the most thorough. Legacy one-shot and agentic modes still work, running as basic and advanced.

- **Zero data retention:** available
- **Details:** https://vercel.com/ai-gateway/models/search

### tako/search

Search Tako's curated knowledge graph and the live web for source-grounded data cards, citations, and visualizations. Use it as a built-in tool through the AI Gateway to give any model access to structured data and current web information.

- **Details:** https://vercel.com/ai-gateway/models/search

### browserbase/search

Search the web with Browserbase's Search API for fast, structured web discovery, returning result titles, URLs, and metadata without opening a browser session. Use it as a built-in tool through the AI Gateway to give any model access to real-time web information.

- **Details:** https://vercel.com/ai-gateway/models/search

### browserbase/fetch

Fetch the contents of a URL with Browserbase, optionally through proxies, returning the upstream response body unchanged or content extracted as Markdown or JSON. Use it as a built-in tool through the AI Gateway to let any model read a page it already has a URL for.

- **Details:** https://vercel.com/ai-gateway/models/fetch


If you encounter incorrect information, can’t find what you need, need a capability Vercel doesn’t currently offer, or have any other feedback to report, please do it by following the instructions at https://vercel.com/feedback.md.