Model catalog
| Model | Modality | Status | Context | Pricing |
|---|---|---|---|---|
| qwen3.6-35b-a3b-awq
Fast general text model |
chat | live | 16,000 | $0.6 / 1m output tokens |
| qwen3.8-27b
Text + image input |
chat | live | 64,000 | $0.9 / 1m output tokens |
| qwen3-vl-30b-a3b
Vision: 128 images or 1 video |
chat | live | 80,000 | $0.9 / 1m output tokens |
| z-image-turbo
~3 s sync generation |
image | live | — | $0.008 / image |
| qwen-image
High quality, async ~5 min |
image | live | — | $0.04 / image |
| qwen-image-edit
Edit with input images, ~55 s |
image | live | — | $0.03 / image |
| minimax-h3
768p 24fps with audio, async ~7 min |
video | live | — | $0.03 / second video |
| kokoro
WAV 24 kHz |
tts | live | — | $0.002 / minute audio |
| qwen3-asr
Text + language detection |
asr | live | — | $0.003 / minute audio |
| whisper-turbo
Timestamped chunks |
asr | live | — | $0.003 / minute audio |
| whisperx
Word alignment + diarization |
asr | live | — | $0.005 / minute audio |
| pyannote-diarize
Speaker segments |
asr | live | — | $0.003 / minute audio |
| sam2
Segmentation masks |
vision | live | — | $0.005 / request |
| talknet-asd
Active speaker detection |
vision | live | — | $0.002 / second video |
| acestep
WAV 48 kHz stereo, async |
music | live | — | $0.03 / minute audio |
| granite
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| gemma
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| nemotron
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| qwen3-coder
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| embeddings
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| reranker
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| chroma
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| chatterbox
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| omnivoice
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| diffrhythm2
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| yue
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| stable-audio-open
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| qwen3-vl-8b
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| glm-ocr
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| qwen2.5-omni
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| longvideoagent
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
| qwen3.6-3090
Scaled to zero; warms on demand |
chat | cold | — | $0 / request |
Cold models warm on first request (503 + Retry-After while starting). Planned models are on the roadmap.