# AIgateway — full model catalog

> Every model routed through AIgateway, with price, context, capabilities, and release date.
> Same endpoint (https://api.aigateway.sh/v1) for all of them. Same auth. OpenAI-compatible request shape.

> Generated 2026-04-24T10:23:38.501Z · catalog size 169

## Comparison matrix — frontier models

| Model | Context | Input / 1M | Output / 1M | Tools | Vision | Reasoning |
|---|---|---|---|---|---|---|
| anthropic/claude-opus-4.7 | 1M | $5 | $25 | yes | yes | yes |
| anthropic/claude-sonnet-4.6 | 200K | $3 | $15 | yes | yes | yes |
| anthropic/claude-haiku-4.5 | 200K | $1 | $5 | yes | yes | no |
| openai/gpt-5.4 | 400K | $2.5 | $15 | yes | yes | yes |
| openai/gpt-5.4-mini | 256K | $0.75 | $4.5 | yes | yes | no |
| google/gemini-3.1-pro | 2M | $2 | $12 | yes | no | no |
| google/gemini-3-flash | 1M | $0.5 | $3 | yes | no | no |
| moonshot/kimi-k2.6 | 256K | $0.95 | $4 | yes | yes | yes |

## Pricing table (all catalog models)

| Slug | Modality | Input / 1M | Output / 1M |
|---|---|---|---|
| anthropic/claude-opus-4.7 | text | $5.00 | $25.00 |
| openai/gpt-5.4 | text | $2.50 | $15.00 |
| google/gemini-3.1-pro | text | $2.00 | $12.00 |
| moonshot/kimi-k2.6 | text | $0.950 | $4.00 |
| anthropic/claude-sonnet-4.6 | text | $3.00 | $15.00 |
| minimax/m2.7 | text | $0.300 | $1.20 |
| openai/gpt-5.4-mini | text | $0.750 | $4.50 |
| xai/grok-4 | text | $5.00 | $15.00 |
| google/gemini-3-flash | text | $0.500 | $3.00 |
| anthropic/claude-haiku-4.5 | text | $1.00 | $5.00 |
| black-forest-labs/flux-2-klein-9b | image | $0.0400/img | — |
| google/imagen-4 | image | $0.0400/img | — |
| google/veo-3.1 | video | $0.400/sec | — |
| deepgram/nova-3 | audio-stt | $0.0052/min | — |
| deepgram/aura-2-en | audio-tts | $0.0300/1K-chars | — |
| baai/bge-m3 | embedding | $0.012 | — |
| meta/llama-4-scout-17b-16e-instruct | text | $0.270 | $0.850 |
| openai/gpt-oss-120b | text | $0.350 | $0.750 |
| qwen/qwen2.5-coder-32b-instruct | text | $0.660 | $1.00 |
| alibaba/qwen3-max | text | $1.20 | $6.00 |
| alibaba/qwen3.5-397b-a17b | text | $0.600 | $3.60 |
| pixverse/v5.6 | video | $0.080/sec | — |
| pixverse/v6 | video | $0.025/sec | — |
| vidu/q3-pro | video | $0.050/sec | — |
| vidu/q3-turbo | video | $0.040/sec | — |
| alibaba/wan-2.6-image | image | $0.0300/img | — |
| openai/gpt-image-1.5 | image | $5.00 | $10.00 |
| runwayml/gen-4.5 | video | $0.120/sec | — |
| minimax/music-2.6 | music | $0.150/sec | — |
| assemblyai/universal-3-pro | audio-stt | $0.0035/min | — |
| anthropic/claude-opus-4.6 | text | $5.00 | $25.00 |
| anthropic/claude-sonnet-4 | text | $3.00 | $15.00 |
| anthropic/claude-sonnet-4.5 | text | $3.00 | $15.00 |
| google/gemini-3.1-flash-lite | text | $0.250 | $1.50 |
| openai/gpt-4.1 | text | $2.00 | $8.00 |
| openai/gpt-4.1-mini | text | $0.400 | $1.60 |
| openai/gpt-5 | text | $1.25 | $10.00 |
| openai/gpt-5.4-nano | text | $0.200 | $1.25 |
| recraft/recraftv4 | image | $0.0400/img | — |
| recraft/recraftv4-pro | image | $0.2500/img | — |
| recraft/recraftv4-pro-vector | image | $0.3000/img | — |
| recraft/recraftv4-vector | image | $0.0800/img | — |
| inworld/tts-1.5-max | audio-tts | $0.0500/1K-chars | — |
| inworld/tts-1.5-mini | audio-tts | $0.0250/1K-chars | — |
| minimax/speech-2.8-hd | audio-tts | $0.1000/1K-chars | — |
| minimax/speech-2.8-turbo | audio-tts | $0.0600/1K-chars | — |
| openai/tts-1 | audio-tts | $0.0150/1K-chars | — |
| minimax/hailuo-2.3 | video | $0.047/sec | — |
| openai/gpt-4o-transcribe | stt | $0.0060/min | — |
| openai/o4-mini | text | $1.10 | $4.40 |
| openai/tts-1-hd | tts | $0.0300/1K-chars | — |
| minimax/hailuo-2.3-fast | video | $0.032/sec | — |
| moonshot/kimi-k2.5 | text | $0.600 | $3.00 |
| bytedance/seedream-4.0 | image | $0.0300/img | — |
| bytedance/seedream-4.5 | image | $0.0400/img | — |
| bytedance/seedream-5-lite | image | $0.0350/img | — |
| google/nano-banana | image | $0.300 | $30.00 |
| google/nano-banana-2 | image | $0.500 | $60.00 |
| google/nano-banana-pro | image | $2.00 | $120.00 |
| google/veo-3 | video | $0.200/sec | — |
| google/veo-3-fast | video | $0.080/sec | — |
| google/veo-3.1-fast | video | $0.080/sec | — |
| google/gemma-4-26b-a4b-it | text | $0.100 | $0.300 |
| cohere/cohere-transcribe-03-2026 | audio-stt | $0.0060/min | — |
| mistral/voxtral-tts-26-03 | audio-tts | $12.0000/1K-chars | — |
| mistral/mistral-small-4-0-26-03 | text | $0.200 | $0.600 |
| nvidia/nemotron-3-120b-a12b | text | $0.500 | $1.50 |
| mistral/voxtral-mini-transcribe-realtime-26-02 | audio-stt | $0.0080/min | — |
| mistral/voxtral-mini-transcribe-26-02 | audio-stt | $0.0060/min | — |
| zai-org/glm-4.7-flash | text | $0.060 | $0.400 |
| black-forest-labs/flux-2-klein-4b | image | $0.0000/img | — |
| black-forest-labs/flux-2-dev | image | $0.0000/img | — |
| cartesia/sonic-3 | audio-tts | $30.0000/1K-chars | — |
| deepgram/aura-2-es | audio-tts | $0.0300/1K-chars | — |
| ibm-granite/granite-4.0-h-micro | text | $0.017 | $0.110 |
| deepgram/flux | audio-stt | $0.0077/min | — |
| pfnet/plamo-embedding-1b | embedding | $0.019 | — |
| aisingapore/gemma-sea-lion-v4-27b-it | text | $0.350 | $0.560 |
| ai4bharat/indictrans2-en-indic-1B | translation | $0.340 | $0.340 |
| xai/grok-4-fast | text | $0.500 | $2.00 |
| google/embeddinggemma-300m | embedding | $0.020 | — |
| deepgram/aura-1 | audio-tts | $0.0150/1K-chars | — |
| leonardo/lucid-origin | image | $0.0039/img | — |
| leonardo/phoenix-1.0 | image | $0.0033/img | — |
| openai/gpt-oss-20b | text | $0.200 | $0.300 |
| pipecat-ai/smart-turn-v2 | classification | $0.0003/min | — |
| qwen/qwen3-embedding-0.6b | embedding | $0.012 | — |
| qwen/qwen3-30b-a3b-fp8 | text | $0.051 | $0.340 |
| google/gemma-3-12b-it | text | $0.350 | $0.560 |
| mistralai/mistral-small-3.1-24b-instruct | text | $0.350 | $0.550 |
| perplexity/sonar-reasoning-pro | text | $2.00 | $8.00 |
| qwen/qwq-32b | reasoning | $0.200 | $0.400 |
| perplexity/sonar-deep-research | text | $2.00 | $8.00 |
| baai/bge-reranker-base | reranking | $0.003 | — |
| deepseek/deepseek-r1-distill-qwen-32b | reasoning | $0.500 | $4.88 |
| meta/llama-guard-3-8b | moderation | $0.480 | $0.030 |
| meta/llama-3.3-70b-instruct-fp8-fast | text | $0.290 | $2.25 |
| meta/llama-3.2-11b-vision-instruct | text | $0.049 | $0.680 |
| meta/llama-3.2-1b-instruct | text | $0.027 | $0.200 |
| meta/llama-3.2-3b-instruct | text | $0.051 | $0.340 |
| black-forest-labs/flux-1-schnell | image | $0.0004/img | — |
| meta/llama-3.1-8b-instruct-awq | text | $0.120 | $0.270 |
| meta/llama-3.1-8b-instruct-fp8 | text | $0.150 | $0.290 |
| meta/llama-3.1-70b-instruct | text | $0.290 | $0.600 |
| meta/llama-3.1-8b-instruct | text | $0.050 | $0.100 |
| meta/llama-3.1-8b-instruct-fast | text | $0.050 | $0.100 |
| myshell-ai/melotts | audio-tts | $0.0002/min | — |
| openai/whisper-large-v3-turbo | audio-stt | $0.0005/min | — |
| meta/llama-3-8b-instruct-awq | text | $0.120 | $0.270 |
| openai/whisper-tiny-en | audio-stt | $0.0005/min | — |
| meta/llama-3-8b-instruct | text | $0.280 | $0.830 |
| hf/meta-llama/meta-llama-3-8b-instruct | text | $0.050 | $0.100 |
| google/gemma-2b-it-lora | text | $0.030 | $0.060 |
| google/gemma-7b-it-lora | text | $0.080 | $0.160 |
| meta-llama/llama-2-7b-chat-hf-lora | text | $0.040 | $0.080 |
| hf/mistral/mistral-7b-instruct-v0.2 | text | $0.050 | $0.100 |
| mistral/mistral-7b-instruct-v0.2-lora | text | $0.050 | $0.100 |
| hf/google/gemma-7b-it | text | $0.080 | $0.160 |
| hf/nousresearch/hermes-2-pro-mistral-7b | text | $0.050 | $0.100 |
| hf/nexusflow/starling-lm-7b-beta | text | $0.050 | $0.100 |
| unum/uform-gen2-qwen-500m | vision | $0.0000/img | — |
| deepseek/deepseek-math-7b-instruct | text | — | — |
| defog/sqlcoder-7b-2 | text | $0.050 | $0.100 |
| microsoft/phi-2 | text | $0.020 | $0.040 |
| bytedance/stable-diffusion-xl-lightning | image | $0.0000/img | — |
| lykon/dreamshaper-8-lcm | image | $0.0008/img | — |
| facebook/bart-large-cnn | text | — | — |
| runwayml/stable-diffusion-v1-5-img2img | image | $0.0000/img | — |
| runwayml/stable-diffusion-v1-5-inpainting | image | $0.0000/img | — |
| qwen/qwen1.5-0.5b-chat | text | $0.010 | $0.020 |
| qwen/qwen1.5-1.8b-chat | text | $0.020 | $0.040 |
| qwen/qwen1.5-14b-chat-awq | text | $0.120 | $0.240 |
| qwen/qwen1.5-7b-chat-awq | text | $0.060 | $0.120 |
| thebloke/discolm-german-7b-v1-awq | text | $0.050 | $0.100 |
| openchat/openchat-3.5-0106 | text | $0.050 | $0.100 |
| tinyllama/tinyllama-1.1b-chat-v1.0 | text | $0.008 | $0.016 |
| hf/thebloke/llamaguard-7b-awq | text | $0.040 | $0.080 |
| fblgit/una-cybertron-7b-v2-bf16 | text | $0.050 | $0.100 |
| hf/thebloke/neural-chat-7b-v3-1-awq | text | $0.050 | $0.100 |
| stabilityai/stable-diffusion-xl-base-1.0 | image | $0.0000/img | — |
| baai/bge-large-en-v1.5 | embedding | $0.204 | — |
| baai/bge-small-en-v1.5 | embedding | $0.020 | — |
| meta/llama-2-7b-chat-fp16 | text | $0.560 | $6.67 |
| mistral/mistral-7b-instruct-v0.1 | text | $0.110 | $0.190 |
| hf/thebloke/deepseek-coder-6.7b-base-awq | text | $0.050 | $0.100 |
| hf/thebloke/deepseek-coder-6.7b-instruct-awq | text | $0.050 | $0.100 |
| hf/thebloke/openhermes-2.5-mistral-7b-awq | text | $0.050 | $0.100 |
| hf/thebloke/zephyr-7b-beta-awq | text | $0.050 | $0.100 |
| llava-hf/llava-1.5-7b-hf | vision | $0.500 | $0.000 |
| hf/thebloke/mistral-7b-instruct-v0.1-awq | text | $0.050 | $0.100 |
| openai/whisper | audio-stt | $0.0005/min | — |
| baai/bge-base-en-v1.5 | embedding | $0.067 | — |
| meta/llama-2-7b-chat-int8 | text | $0.040 | $0.080 |
| microsoft/resnet-50 | classification | $0.0000/img | — |
| huggingface/distilbert-sst-2-int8 | classification | $0.026 | — |
| meta/m2m100-1.2b | translation | $0.340 | $0.340 |
| hf/thebloke/llama-2-13b-chat-awq | text | $0.070 | $0.140 |
| tiiuae/falcon-7b-instruct | text | $0.050 | $0.100 |
| facebook/detr-resnet-50 | object-detection | $0.0000/img | — |
| ai4bharat/indictrans2-en-indic-1B | text | $0.021 | $0.042 |
| baai/bge-reranker-base | rerank | — | — |
| facebook/bart-large-cnn | text | $0.050 | $0.100 |
| facebook/detr-resnet-50 | image | $0.0002/img | — |
| huggingface/distilbert-sst-2-int8 | text | — | — |
| meta/m2m100-1.2b | text | $0.021 | $0.042 |
| microsoft/resnet-50 | image | $0.0002/img | — |
| minimax/music-2.6 | audio-tts | $0.052/sec | — |
| mistral/mistral-moderation-26-03 | moderation | — | — |
| pipecat-ai/smart-turn-v2 | audio-stt | $0.0021/min | — |

## Start here (top picks across modalities)

If you're an agent picking a default, use this ordered list. Top-1 is the model we'd pick for a general-purpose, quality-first task; further down trades quality for cost or speed.

- anthropic/claude-opus-4.7  [text]  Claude Opus 4.7 · Anthropic released=2026-04-16 ctx=1000000 max_out=128000 caps=[tools,streaming,vision,json,caching,reasoning]  in=$5.00/Mtok out=$25.00/Mtok cache-read=$0.500/Mtok
- openai/gpt-5.4  [text]  GPT-5.4 · OpenAI released=2026-04-08 ctx=128000 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning,batch]  in=$2.50/Mtok out=$15.00/Mtok cache-read=$0.250/Mtok
- google/gemini-3.1-pro  [text]  Gemini 3.1 Pro · Google released=2026-04-13 ctx=1000000 max_out=65536 caps=[tools,streaming,vision,json,caching,reasoning]  in=$2.00/Mtok out=$12.00/Mtok cache-read=$0.200/Mtok
- moonshot/kimi-k2.6  [text]  Kimi K2.6 · Moonshot released=2026-04-20 ctx=262144 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning]  in=$0.950/Mtok out=$4.00/Mtok cache-read=$0.160/Mtok
- anthropic/claude-sonnet-4.6  [text]  Claude Sonnet 4.6 · Anthropic released=2026-04-13 ctx=200000 max_out=128000 caps=[tools,streaming,vision,json,caching,reasoning]  in=$3.00/Mtok out=$15.00/Mtok cache-read=$0.300/Mtok
- minimax/m2.7  [text]  M2.7 · MiniMax released=2026-04-13 ctx=128000 max_out=4096 caps=[streaming]  in=$0.300/Mtok out=$1.20/Mtok
- openai/gpt-5.4-mini  [text]  GPT-5.4 Mini · OpenAI released=2026-04-13 ctx=128000 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning,batch]  in=$0.750/Mtok out=$4.50/Mtok cache-read=$0.075/Mtok
- xai/grok-4  [text]  Grok 4 · xAI released=2025-07-09 ctx=256000 max_out=16384 caps=[tools,streaming,vision,json,reasoning]  in=$5.00/Mtok out=$15.00/Mtok
- google/gemini-3-flash  [text]  Gemini 3 Flash · Google released=2026-04-13 ctx=1000000 max_out=8192 caps=[tools,streaming,vision,json,caching,reasoning]  in=$0.500/Mtok out=$3.00/Mtok cache-read=$0.050/Mtok
- anthropic/claude-haiku-4.5  [text]  Claude Haiku 4.5 · Anthropic released=2026-04-13 ctx=200000 max_out=8192 caps=[tools,streaming,vision,json,caching,reasoning]  in=$1.00/Mtok out=$5.00/Mtok cache-read=$0.100/Mtok

## Full catalog — by modality

### text (86)

- anthropic/claude-opus-4.7  [text]  Claude Opus 4.7 · Anthropic released=2026-04-16 ctx=1000000 max_out=128000 caps=[tools,streaming,vision,json,caching,reasoning]  in=$5.00/Mtok out=$25.00/Mtok cache-read=$0.500/Mtok
- openai/gpt-5.4  [text]  GPT-5.4 · OpenAI released=2026-04-08 ctx=128000 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning,batch]  in=$2.50/Mtok out=$15.00/Mtok cache-read=$0.250/Mtok
- google/gemini-3.1-pro  [text]  Gemini 3.1 Pro · Google released=2026-04-13 ctx=1000000 max_out=65536 caps=[tools,streaming,vision,json,caching,reasoning]  in=$2.00/Mtok out=$12.00/Mtok cache-read=$0.200/Mtok
- moonshot/kimi-k2.6  [text]  Kimi K2.6 · Moonshot released=2026-04-20 ctx=262144 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning]  in=$0.950/Mtok out=$4.00/Mtok cache-read=$0.160/Mtok
- anthropic/claude-sonnet-4.6  [text]  Claude Sonnet 4.6 · Anthropic released=2026-04-13 ctx=200000 max_out=128000 caps=[tools,streaming,vision,json,caching,reasoning]  in=$3.00/Mtok out=$15.00/Mtok cache-read=$0.300/Mtok
- minimax/m2.7  [text]  M2.7 · MiniMax released=2026-04-13 ctx=128000 max_out=4096 caps=[streaming]  in=$0.300/Mtok out=$1.20/Mtok
- openai/gpt-5.4-mini  [text]  GPT-5.4 Mini · OpenAI released=2026-04-13 ctx=128000 max_out=16384 caps=[tools,streaming,vision,json,caching,reasoning,batch]  in=$0.750/Mtok out=$4.50/Mtok cache-read=$0.075/Mtok
- xai/grok-4  [text]  Grok 4 · xAI released=2025-07-09 ctx=256000 max_out=16384 caps=[tools,streaming,vision,json,reasoning]  in=$5.00/Mtok out=$15.00/Mtok
- google/gemini-3-flash  [text]  Gemini 3 Flash · Google released=2026-04-13 ctx=1000000 max_out=8192 caps=[tools,streaming,vision,json,caching,reasoning]  in=$0.500/Mtok out=$3.00/Mtok cache-read=$0.050/Mtok
- anthropic/claude-haiku-4.5  [text]  Claude Haiku 4.5 · Anthropic released=2026-04-13 ctx=200000 max_out=8192 caps=[tools,streaming,vision,json,caching,reasoning]  in=$1.00/Mtok out=$5.00/Mtok cache-read=$0.100/Mtok
- meta/llama-4-scout-17b-16e-instruct  [text]  Llama-4-Scout-17b-16e-Instruct · Meta released=2025-04-05 ctx=131000 max_out=4096 caps=[tools,streaming,vision,json]  in=$0.270/Mtok out=$0.850/Mtok
- openai/gpt-oss-120b  [text]  Gpt-Oss-120b · OpenAI released=2025-08-05 ctx=128000 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.350/Mtok out=$0.750/Mtok
- qwen/qwen2.5-coder-32b-instruct  [text]  Qwen2.5-Coder-32b-Instruct · Alibaba Qwen released=2025-02-27 ctx=32768 max_out=4096 caps=[tools,streaming,json]  in=$0.660/Mtok out=$1.00/Mtok
- alibaba/qwen3-max  [text]  Qwen 3 Max · Alibaba released=2026-04-15 ctx=262144 max_out=4096 caps=[streaming,reasoning]  in=$1.20/Mtok out=$6.00/Mtok
- alibaba/qwen3.5-397b-a17b  [text]  Qwen 3.5 397B A17B · Alibaba released=2026-04-15 ctx=262144 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.600/Mtok out=$3.60/Mtok
- anthropic/claude-opus-4.6  [text]  Claude Opus 4.6 · Anthropic released=2026-04-13 ctx=1000000 max_out=128000 caps=[tools,streaming,vision,json,caching,reasoning]  in=$5.00/Mtok out=$25.00/Mtok cache-read=$0.500/Mtok
- anthropic/claude-sonnet-4  [text]  Claude Sonnet 4 · Anthropic released=2026-04-13 ctx=200000 max_out=16000 caps=[tools,streaming,json,reasoning]  in=$3.00/Mtok out=$15.00/Mtok
- anthropic/claude-sonnet-4.5  [text]  Claude Sonnet 4.5 · Anthropic released=2026-04-13 ctx=200000 max_out=8192 caps=[tools,streaming,json,reasoning]  in=$3.00/Mtok out=$15.00/Mtok
- google/gemini-3.1-flash-lite  [text]  Gemini 3.1 Flash Lite · Google released=2026-04-13 ctx=1000000 max_out=8192 caps=[tools,streaming,vision,json,caching]  in=$0.250/Mtok out=$1.50/Mtok cache-read=$0.025/Mtok
- openai/gpt-4.1  [text]  GPT-4.1 · OpenAI released=2026-04-13 ctx=1047576 max_out=32768 caps=[tools,streaming,json,reasoning]  in=$2.00/Mtok out=$8.00/Mtok
- openai/gpt-4.1-mini  [text]  GPT-4.1 Mini · OpenAI released=2026-04-13 ctx=1047576 max_out=32768 caps=[tools,streaming,json]  in=$0.400/Mtok out=$1.60/Mtok
- openai/gpt-5  [text]  GPT-5 · OpenAI released=2026-04-13 ctx=128000 max_out=16384 caps=[tools,streaming,json,reasoning]  in=$1.25/Mtok out=$10.00/Mtok
- openai/gpt-5.4-nano  [text]  GPT-5.4 Nano · OpenAI released=2026-04-13 ctx=128000 max_out=16384 caps=[tools,streaming,json,reasoning,batch]  in=$0.200/Mtok out=$1.25/Mtok cache-read=$0.020/Mtok
- openai/o4-mini  [text]  o4-mini · openai released=2026-04-13 ctx=200000 max_out=100000 caps=[streaming,reasoning]  in=$1.10/Mtok out=$4.40/Mtok
- moonshot/kimi-k2.5  [text]  Kimi-K2.5 · Moonshot released=2026-04-08 ctx=128000 max_out=8192 caps=[tools,streaming,vision,json,reasoning]  in=$0.600/Mtok out=$3.00/Mtok
- google/gemma-4-26b-a4b-it  [text]  Gemma-4-26b-A4b-IT · Google released=2026-04-02 ctx=256000 max_out=4096 caps=[tools,streaming,vision,json,reasoning]  in=$0.100/Mtok out=$0.300/Mtok
- mistral/mistral-small-4-0-26-03  [text]  Mistral Small 4 · Mistral released=2026-03-01 ctx=131072 max_out=16384 caps=[tools,streaming,vision,json]  in=$0.200/Mtok out=$0.600/Mtok
- nvidia/nemotron-3-120b-a12b  [text]  Nemotron-3-120b-A12b · NVIDIA released=2026-02-24 ctx=256000 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.500/Mtok out=$1.50/Mtok
- zai-org/glm-4.7-flash  [text]  Glm-4.7-Flash · Zhipu AI released=2026-01-28 ctx=131072 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.060/Mtok out=$0.400/Mtok
- ibm-granite/granite-4.0-h-micro  [text]  Granite-4.0-H-Micro · IBM released=2025-10-07 ctx=131000 max_out=4096 caps=[tools,streaming,json]  in=$0.017/Mtok out=$0.110/Mtok
- aisingapore/gemma-sea-lion-v4-27b-it  [text]  Gemma-Sea-Lion-V4-27b-IT · AI Singapore released=2025-09-23 ctx=128000 max_out=4096 caps=[streaming]  in=$0.350/Mtok out=$0.560/Mtok
- xai/grok-4-fast  [text]  Grok 4 Fast · xAI released=2025-09-19 ctx=256000 max_out=16384 caps=[tools,streaming,json]  in=$0.500/Mtok out=$2.00/Mtok
- openai/gpt-oss-20b  [text]  Gpt-Oss-20b · OpenAI released=2025-08-05 ctx=128000 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.200/Mtok out=$0.300/Mtok
- qwen/qwen3-30b-a3b-fp8  [text]  Qwen3-30b-A3b-Fp8 · Alibaba Qwen released=2025-04-30 ctx=32768 max_out=4096 caps=[tools,streaming,json,reasoning]  in=$0.051/Mtok out=$0.340/Mtok
- google/gemma-3-12b-it  [text]  Gemma-3-12b-IT · Google released=2025-03-18 ctx=80000 max_out=4096 caps=[streaming,vision]  in=$0.350/Mtok out=$0.560/Mtok
- mistralai/mistral-small-3.1-24b-instruct  [text]  Mistral-Small-3.1-24b-Instruct · Mistral released=2025-03-18 ctx=128000 max_out=4096 caps=[tools,streaming,vision,json]  in=$0.350/Mtok out=$0.550/Mtok
- perplexity/sonar-reasoning-pro  [text]  Sonar Reasoning Pro · Perplexity released=2025-03-07 ctx=127000 max_out=8192 caps=[streaming,reasoning]  in=$2.00/Mtok out=$8.00/Mtok
- perplexity/sonar-deep-research  [text]  Sonar Deep Research · Perplexity released=2025-02-14 ctx=127000 max_out=16384 caps=[streaming,reasoning]  in=$2.00/Mtok out=$8.00/Mtok
- meta/llama-3.3-70b-instruct-fp8-fast  [text]  Llama-3.3-70b-Instruct-Fp8-Fast · Meta released=2024-12-06 ctx=24000 max_out=8192 caps=[tools,streaming,json]  in=$0.290/Mtok out=$2.25/Mtok
- meta/llama-3.2-11b-vision-instruct  [text]  Llama-3.2-11b-Vision-Instruct · Meta released=2024-09-25 ctx=128000 max_out=4096 caps=[streaming,vision]  in=$0.049/Mtok out=$0.680/Mtok
- meta/llama-3.2-1b-instruct  [text]  Llama-3.2-1b-Instruct · Meta released=2024-09-25 ctx=60000 max_out=4096 caps=[streaming]  in=$0.027/Mtok out=$0.200/Mtok
- meta/llama-3.2-3b-instruct  [text]  Llama-3.2-3b-Instruct · Meta released=2024-09-25 ctx=80000 max_out=4096 caps=[streaming]  in=$0.051/Mtok out=$0.340/Mtok
- meta/llama-3.1-8b-instruct-awq  [text]  Llama-3.1-8b-Instruct-Awq · Meta released=2024-07-25 ctx=8192 max_out=4096 caps=[streaming]  in=$0.120/Mtok out=$0.270/Mtok
- meta/llama-3.1-8b-instruct-fp8  [text]  Llama-3.1-8b-Instruct-Fp8 · Meta released=2024-07-25 ctx=32000 max_out=4096 caps=[streaming]  in=$0.150/Mtok out=$0.290/Mtok
- meta/llama-3.1-70b-instruct  [text]  Llama-3.1-70b-Instruct · Meta released=2024-07-23 ctx=131072 max_out=8192 caps=[streaming]  in=$0.290/Mtok out=$0.600/Mtok
- meta/llama-3.1-8b-instruct  [text]  Llama-3.1-8b-Instruct · Meta released=2024-07-23 ctx=131072 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- meta/llama-3.1-8b-instruct-fast  [text]  Llama-3.1-8b-Instruct-Fast · Meta released=2024-07-23 ctx=131072 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- meta/llama-3-8b-instruct-awq  [text]  Llama-3-8b-Instruct-Awq · Meta released=2024-05-09 ctx=8192 max_out=4096 caps=[streaming]  in=$0.120/Mtok out=$0.270/Mtok
- meta/llama-3-8b-instruct  [text]  Llama-3-8b-Instruct · Meta released=2024-04-18 ctx=7968 max_out=4096 caps=[streaming]  in=$0.280/Mtok out=$0.830/Mtok
- hf/meta-llama/meta-llama-3-8b-instruct  [text]  Meta-Llama-3-8b-Instruct · Hugging Face released=2024-04-18 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- google/gemma-2b-it-lora  [text]  Gemma-2b-IT-Lora · Google released=2024-04-02 ctx=8192 max_out=4096 caps=[streaming]  in=$0.030/Mtok out=$0.060/Mtok
- google/gemma-7b-it-lora  [text]  Gemma-7b-IT-Lora · Google released=2024-04-02 ctx=3500 max_out=4096 caps=[streaming]  in=$0.080/Mtok out=$0.160/Mtok
- meta-llama/llama-2-7b-chat-hf-lora  [text]  Llama-2-7b-Chat-HF-Lora · Meta released=2024-04-02 ctx=8192 max_out=4096 caps=[streaming]  in=$0.040/Mtok out=$0.080/Mtok
- hf/mistral/mistral-7b-instruct-v0.2  [text]  Mistral-7b-Instruct-V0.2 · Hugging Face released=2024-04-02 ctx=3072 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- mistral/mistral-7b-instruct-v0.2-lora  [text]  Mistral-7b-Instruct-V0.2-Lora · Mistral released=2024-04-01 ctx=15000 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/google/gemma-7b-it  [text]  Gemma-7b-IT · Hugging Face released=2024-04-01 ctx=8192 max_out=4096 caps=[streaming]  in=$0.080/Mtok out=$0.160/Mtok
- hf/nousresearch/hermes-2-pro-mistral-7b  [text]  Hermes-2-Pro-Mistral-7b · Hugging Face released=2024-04-01 ctx=24000 max_out=4096 caps=[tools,streaming,json]  in=$0.050/Mtok out=$0.100/Mtok
- hf/nexusflow/starling-lm-7b-beta  [text]  Starling-LM-7b-Beta · Hugging Face released=2024-03-19 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- deepseek/deepseek-math-7b-instruct  [text]  Deepseek-Math-7b-Instruct · DeepSeek released=2024-02-27 ctx=4096 max_out=4096 caps=[streaming]  
- defog/sqlcoder-7b-2  [text]  Sqlcoder-7b-2 · Defog released=2024-02-27 ctx=10000 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- microsoft/phi-2  [text]  Phi-2 · Microsoft released=2024-02-27 ctx=2048 max_out=4096 caps=[streaming]  in=$0.020/Mtok out=$0.040/Mtok
- facebook/bart-large-cnn  [text]  Bart-Large-CNN · Meta released=2024-02-27  
- qwen/qwen1.5-0.5b-chat  [text]  Qwen1.5-0.5b-Chat · Alibaba Qwen released=2024-02-05 ctx=4096 max_out=4096 caps=[streaming]  in=$0.010/Mtok out=$0.020/Mtok
- qwen/qwen1.5-1.8b-chat  [text]  Qwen1.5-1.8b-Chat · Alibaba Qwen released=2024-02-05 ctx=4096 max_out=4096 caps=[streaming]  in=$0.020/Mtok out=$0.040/Mtok
- qwen/qwen1.5-14b-chat-awq  [text]  Qwen1.5-14b-Chat-Awq · Alibaba Qwen released=2024-02-05 ctx=4096 max_out=4096 caps=[streaming]  in=$0.120/Mtok out=$0.240/Mtok
- qwen/qwen1.5-7b-chat-awq  [text]  Qwen1.5-7b-Chat-Awq · Alibaba Qwen released=2024-02-05 ctx=4096 max_out=4096 caps=[streaming]  in=$0.060/Mtok out=$0.120/Mtok
- thebloke/discolm-german-7b-v1-awq  [text]  Discolm-German-7b-V1-Awq · TheBloke released=2024-01-24 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- openchat/openchat-3.5-0106  [text]  Openchat-3.5-0106 · OpenChat released=2024-01-06 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- tinyllama/tinyllama-1.1b-chat-v1.0  [text]  Tinyllama-1.1b-Chat-V1.0 · TinyLlama released=2024-01-04 ctx=2048 max_out=4096 caps=[streaming]  in=$0.008/Mtok out=$0.016/Mtok
- hf/thebloke/llamaguard-7b-awq  [text]  Llamaguard-7b-Awq · Hugging Face released=2023-12-11 ctx=4096 max_out=4096 caps=[streaming]  in=$0.040/Mtok out=$0.080/Mtok
- fblgit/una-cybertron-7b-v2-bf16  [text]  Una-Cybertron-7b-V2-Bf16 · FBL released=2023-12-01 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/thebloke/neural-chat-7b-v3-1-awq  [text]  Neural-Chat-7b-V3-1-Awq · Hugging Face released=2023-11-14 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- meta/llama-2-7b-chat-fp16  [text]  Llama-2-7b-Chat-Fp16 · Meta released=2023-11-07 ctx=4096 max_out=4096 caps=[streaming]  in=$0.560/Mtok out=$6.67/Mtok
- mistral/mistral-7b-instruct-v0.1  [text]  Mistral-7b-Instruct-V0.1 · Mistral released=2023-11-07 ctx=2824 max_out=4096 caps=[streaming]  in=$0.110/Mtok out=$0.190/Mtok
- hf/thebloke/deepseek-coder-6.7b-base-awq  [text]  Deepseek-Coder-6.7b-Base-Awq · Hugging Face released=2023-11-03 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/thebloke/deepseek-coder-6.7b-instruct-awq  [text]  Deepseek-Coder-6.7b-Instruct-Awq · Hugging Face released=2023-11-03 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/thebloke/openhermes-2.5-mistral-7b-awq  [text]  Openhermes-2.5-Mistral-7b-Awq · Hugging Face released=2023-11-02 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/thebloke/zephyr-7b-beta-awq  [text]  Zephyr-7b-Beta-Awq · Hugging Face released=2023-10-27 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- hf/thebloke/mistral-7b-instruct-v0.1-awq  [text]  Mistral-7b-Instruct-V0.1-Awq · Hugging Face released=2023-09-27 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- meta/llama-2-7b-chat-int8  [text]  Llama-2-7b-Chat-Int8 · Meta released=2023-09-25 ctx=8192 max_out=4096 caps=[streaming]  in=$0.040/Mtok out=$0.080/Mtok
- hf/thebloke/llama-2-13b-chat-awq  [text]  Llama-2-13b-Chat-Awq · Hugging Face released=2023-07-18 ctx=4096 max_out=4096 caps=[streaming]  in=$0.070/Mtok out=$0.140/Mtok
- tiiuae/falcon-7b-instruct  [text]  Falcon-7b-Instruct · TII released=2023-05-25 ctx=4096 max_out=4096 caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- ai4bharat/indictrans2-en-indic-1B  [text]  IndicTrans2 EN→Indic 1B · AI4Bharat  in=$0.021/Mtok out=$0.042/Mtok
- facebook/bart-large-cnn  [text]  BART Large CNN · Meta caps=[streaming]  in=$0.050/Mtok out=$0.100/Mtok
- huggingface/distilbert-sst-2-int8  [text]  DistilBERT SST-2 · Hugging Face  $0.00000/req
- meta/m2m100-1.2b  [text]  M2M100 1.2B · Meta  in=$0.021/Mtok out=$0.042/Mtok

### reasoning (2)

- qwen/qwq-32b  [reasoning]  Qwq-32b · Alibaba Qwen released=2025-03-05 ctx=24000 max_out=4096 caps=[streaming,json,reasoning]  in=$0.200/Mtok out=$0.400/Mtok
- deepseek/deepseek-r1-distill-qwen-32b  [reasoning]  Deepseek-R1-Distill-Qwen-32b · DeepSeek released=2025-01-22 ctx=80000 max_out=4096 caps=[streaming,json,reasoning]  in=$0.500/Mtok out=$4.88/Mtok

### vision (2)

- unum/uform-gen2-qwen-500m  [vision]  Uform-Gen2-Qwen-500m · Unum released=2024-02-27 ctx=4096 max_out=4096 caps=[streaming,vision]  $0.0000/img
- llava-hf/llava-1.5-7b-hf  [vision]  Llava-1.5-7b-HF · LLaVA released=2023-10-05 ctx=4096 max_out=4096 caps=[streaming,vision]  in=$0.500/Mtok out=$0.000/Mtok

### image (26)

- black-forest-labs/flux-2-klein-9b  [image]  Flux-2-Klein-9b · Black Forest Labs released=2026-01-14  $0.0400/img
- google/imagen-4  [image]  Imagen 4 · Google released=2026-04-14  $0.0400/img
- alibaba/wan-2.6-image  [image]  Wan 2.6 Image · Alibaba released=2026-04-14  $0.0300/img
- openai/gpt-image-1.5  [image]  GPT Image 1.5 · OpenAI released=2026-04-14  in=$5.00/Mtok out=$10.00/Mtok
- recraft/recraftv4  [image]  Recraft V4 · Recraft released=2026-04-13  $0.0400/img
- recraft/recraftv4-pro  [image]  Recraft V4 Pro · Recraft released=2026-04-13  $0.2500/img
- recraft/recraftv4-pro-vector  [image]  Recraft V4 Pro Vector · Recraft released=2026-04-13  $0.3000/img
- recraft/recraftv4-vector  [image]  Recraft V4 Vector · Recraft released=2026-04-13  $0.0800/img
- bytedance/seedream-4.0  [image]  Seedream 4.0 · ByteDance released=2026-04-08  $0.0300/img
- bytedance/seedream-4.5  [image]  Seedream 4.5 · ByteDance released=2026-04-08  $0.0400/img
- bytedance/seedream-5-lite  [image]  Seedream 5 Lite · ByteDance released=2026-04-08  $0.0350/img
- google/nano-banana  [image]  Nano Banana · Google released=2026-04-08  in=$0.300/Mtok out=$30.00/Mtok
- google/nano-banana-2  [image]  Nano Banana 2 · Google released=2026-04-08  in=$0.500/Mtok out=$60.00/Mtok
- google/nano-banana-pro  [image]  Nano Banana Pro · Google released=2026-04-08  in=$2.00/Mtok out=$120.00/Mtok $0.1340/img
- black-forest-labs/flux-2-klein-4b  [image]  Flux-2-Klein-4b · Black Forest Labs released=2026-01-14  $0.0000/img
- black-forest-labs/flux-2-dev  [image]  Flux-2-DEV · Black Forest Labs released=2025-11-24  $0.0000/img
- leonardo/lucid-origin  [image]  Lucid-Origin · Leonardo.AI released=2025-08-25  $0.0039/img
- leonardo/phoenix-1.0  [image]  Phoenix-1.0 · Leonardo.AI released=2025-08-25  $0.0033/img
- black-forest-labs/flux-1-schnell  [image]  Flux-1-Schnell · Black Forest Labs released=2024-08-29  $0.0004/img
- bytedance/stable-diffusion-xl-lightning  [image]  Stable-Diffusion-XL-Lightning · ByteDance released=2024-02-27  $0.0000/img
- lykon/dreamshaper-8-lcm  [image]  Dreamshaper-8-Lcm · Lykon released=2024-02-27  $0.0008/img
- runwayml/stable-diffusion-v1-5-img2img  [image]  Stable-Diffusion-V1-5-Img2img · Runway released=2024-02-27  $0.0000/img
- runwayml/stable-diffusion-v1-5-inpainting  [image]  Stable-Diffusion-V1-5-Inpainting · Runway released=2024-02-27  $0.0000/img
- stabilityai/stable-diffusion-xl-base-1.0  [image]  Stable-Diffusion-XL-Base-1.0 · Stability AI released=2023-11-10  $0.0000/img
- facebook/detr-resnet-50  [image]  DETR ResNet-50 · Meta  $0.0002/img
- microsoft/resnet-50  [image]  ResNet-50 · Microsoft  $0.0002/img

### video (11)

- google/veo-3.1  [video]  Veo 3.1 · Google released=2026-04-08  $0.400/sec
- pixverse/v5.6  [video]  Pixverse V5.6 · PixVerse released=2026-04-15  $0.080/sec
- pixverse/v6  [video]  Pixverse V6 · PixVerse released=2026-04-15  $0.025/sec
- vidu/q3-pro  [video]  Vidu Q3 Pro · Vidu released=2026-04-15  $0.050/sec
- vidu/q3-turbo  [video]  Vidu Q3 Turbo · Vidu released=2026-04-15  $0.040/sec
- runwayml/gen-4.5  [video]  RunwayML Gen-4.5 · Runway released=2026-04-14  $0.120/sec
- minimax/hailuo-2.3  [video]  Hailuo 2.3 · MiniMax released=2026-04-13  $0.047/sec
- minimax/hailuo-2.3-fast  [video]  hailuo-2.3-fast · minimax released=2026-04-13  $0.032/sec
- google/veo-3  [video]  Veo 3 · Google released=2026-04-08  $0.200/sec
- google/veo-3-fast  [video]  Veo 3 Fast · Google released=2026-04-08  $0.080/sec
- google/veo-3.1-fast  [video]  Veo 3.1 Fast · Google released=2026-04-08  $0.080/sec

### audio-stt (10)

- deepgram/nova-3  [audio-stt]  Nova-3 · Deepgram released=2025-06-05 caps=[streaming]  $0.0052/min
- assemblyai/universal-3-pro  [audio-stt]  Universal 3 Pro · AssemblyAI released=2026-04-13 caps=[streaming]  $0.0035/min
- cohere/cohere-transcribe-03-2026  [audio-stt]  Cohere Transcribe · Cohere released=2026-03-26  $0.0060/min
- mistral/voxtral-mini-transcribe-realtime-26-02  [audio-stt]  Voxtral Mini Transcribe Realtime · Mistral released=2026-02-01  $0.0080/min
- mistral/voxtral-mini-transcribe-26-02  [audio-stt]  Voxtral Mini Transcribe · Mistral released=2026-02-01  $0.0060/min
- deepgram/flux  [audio-stt]  Flux · Deepgram released=2025-09-29 caps=[streaming]  $0.0077/min
- openai/whisper-large-v3-turbo  [audio-stt]  Whisper-Large-V3-Turbo · OpenAI released=2024-05-22 caps=[streaming]  $0.0005/min
- openai/whisper-tiny-en  [audio-stt]  Whisper-Tiny-EN · OpenAI released=2024-04-22 caps=[streaming]  $0.0005/min
- openai/whisper  [audio-stt]  Whisper · OpenAI released=2023-09-25 caps=[streaming]  $0.0005/min
- pipecat-ai/smart-turn-v2  [audio-stt]  Smart Turn v2 · Pipecat  $0.0021/min

### audio-tts (12)

- deepgram/aura-2-en  [audio-tts]  Aura-2-EN · Deepgram released=2025-10-09 caps=[streaming]  $0.0300/1K-chars
- inworld/tts-1.5-max  [audio-tts]  TTS 1.5 Max · Inworld released=2026-04-13  $0.0500/1K-chars
- inworld/tts-1.5-mini  [audio-tts]  TTS 1.5 Mini · Inworld released=2026-04-13  $0.0250/1K-chars
- minimax/speech-2.8-hd  [audio-tts]  Speech 2.8 HD · MiniMax released=2026-04-13  $0.1000/1K-chars
- minimax/speech-2.8-turbo  [audio-tts]  Speech 2.8 Turbo · MiniMax released=2026-04-13  $0.0600/1K-chars
- openai/tts-1  [audio-tts]  TTS-1 · OpenAI released=2026-04-13  $0.0150/1K-chars
- mistral/voxtral-tts-26-03  [audio-tts]  Voxtral TTS · Mistral released=2026-03-26  $12.0000/1K-chars
- cartesia/sonic-3  [audio-tts]  Sonic 3 · Cartesia released=2025-10-29  $30.0000/1K-chars
- deepgram/aura-2-es  [audio-tts]  Aura-2-ES · Deepgram released=2025-10-09  $0.0300/1K-chars
- deepgram/aura-1  [audio-tts]  Aura-1 · Deepgram released=2025-08-27  $0.0150/1K-chars
- myshell-ai/melotts  [audio-tts]  Melotts · MyShell released=2024-07-19  $0.0002/min
- minimax/music-2.6  [audio-tts]  MiniMax Music 2.6 · MiniMax  $0.052/sec

### embedding (7)

- baai/bge-m3  [embedding]  Bge-M3 · BAAI released=2024-05-22 ctx=60000  in=$0.012/Mtok
- pfnet/plamo-embedding-1b  [embedding]  Plamo-Embedding-1b · PFN released=2025-09-24  in=$0.019/Mtok
- google/embeddinggemma-300m  [embedding]  Embeddinggemma-300m · Google released=2025-09-04 caps=[vision]  in=$0.020/Mtok
- qwen/qwen3-embedding-0.6b  [embedding]  Qwen3-Embedding-0.6b · Alibaba Qwen released=2025-06-18 ctx=8192  in=$0.012/Mtok
- baai/bge-large-en-v1.5  [embedding]  Bge-Large-EN-V1.5 · BAAI released=2023-11-07  in=$0.204/Mtok
- baai/bge-small-en-v1.5  [embedding]  Bge-Small-EN-V1.5 · BAAI released=2023-11-07  in=$0.020/Mtok
- baai/bge-base-en-v1.5  [embedding]  Bge-Base-EN-V1.5 · BAAI released=2023-09-25 ctx=153600  in=$0.067/Mtok

### rerank (1)

- baai/bge-reranker-base  [rerank]  BGE Reranker Base · BAAI  $0.00000/req

### moderation (2)

- meta/llama-guard-3-8b  [moderation]  Llama-Guard-3-8b · Meta released=2025-01-22 ctx=131072 max_out=4096 caps=[streaming]  in=$0.480/Mtok out=$0.030/Mtok
- mistral/mistral-moderation-26-03  [moderation]  Mistral Moderation 2 · Mistral  $0.00010/req

### music (1)

- minimax/music-2.6  [music]  music-2.6 · minimax released=2026-04-14  $0.150/sec

### stt (1)

- openai/gpt-4o-transcribe  [stt]  gpt-4o-transcribe · openai released=2026-04-13 caps=[streaming]  $0.0060/min

### tts (1)

- openai/tts-1-hd  [tts]  tts-1-hd · openai released=2026-04-13  $0.0300/1K-chars

### translation (2)

- ai4bharat/indictrans2-en-indic-1B  [translation]  Indictrans2-EN-Indic-1B · AI4Bharat released=2025-09-23  in=$0.340/Mtok out=$0.340/Mtok
- meta/m2m100-1.2b  [translation]  M2m100-1.2b · Meta released=2023-09-25  in=$0.340/Mtok out=$0.340/Mtok

### classification (3)

- pipecat-ai/smart-turn-v2  [classification]  Smart-Turn-V2 · Pipecat released=2025-08-04  $0.0003/min
- microsoft/resnet-50  [classification]  Resnet-50 · Microsoft released=2023-09-25  $0.0000/img
- huggingface/distilbert-sst-2-int8  [classification]  Distilbert-SST-2-Int8 · Hugging Face released=2023-09-25  in=$0.026/Mtok

### reranking (1)

- baai/bge-reranker-base  [reranking]  BGE-Reranker-Base · BAAI released=2025-02-14  in=$0.003/Mtok

### object-detection (1)

- facebook/detr-resnet-50  [object-detection]  detr-resnet-50 · facebook  $0.0000/img

## How to read a line

Format: `- <provider>/<model>  [<modality>]  <display name> · <provider>  [released=<date>] [ctx=<tokens>] [max_out=<tokens>]  [caps=[...]]  <pricing>`

Pricing units:
- `in=$X/Mtok` / `out=$X/Mtok`: per 1M tokens
- `cache-read=$X/Mtok`: cache hits bill at ~10% of uncached
- `$X/img`: per image (image models)
- `$X/sec`: per video-second (video + music models)
- `$X/min`: per audio minute (STT)
- `$X/1K-chars`: per 1,000 characters (TTS)

## Agent decision rules (recommended)

1. Hard reasoning / code agents / complex tool chains → `anthropic/claude-opus-4.7`
2. General chat, RAG, production default → `anthropic/claude-sonnet-4.6` or `openai/gpt-5.4`
3. Very long context (>500K tokens) → `google/gemini-3.1-pro` (2M) or `openai/gpt-5.4` (1M)
4. Cheap/bulk work → `openai/gpt-5.4-mini`, `google/gemini-3.1-flash-lite`, or `moonshot/kimi-k2.6` (free tier)
5. Image gen → `black-forest-labs/flux-2-klein-9b` (SOTA open) or `google/imagen-4`
6. Video → `google/veo-3.1`
7. STT → `deepgram/nova-3`  ·  TTS → `deepgram/aura-2-en`
8. Embeddings → `baai/bge-m3`  ·  Rerank → `baai/bge-reranker-base`
9. Moderation → `meta/llama-guard-3-8b`

See /llms.txt for the editorial overview and /reference for full API docs.