Qwen 3.5 397B A17B
by Deep Infra · family qwen · listed Feb 2026RTVW
Qwen vision-language model for visual reasoning, documents, and agent tasks
Current rates USD per 1M tokens
Input$0.45
Output$3.00
Cache read$0.22
Cache write—
Against the market
On the record
- Model id
Qwen/Qwen3.5-397B-A17B- Context window
- 262K tokens
- Max output
- 82K tokens
- Input modalities
- text, image, video
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- Jan 2025
- Released
- 2026-02-01
- Last updated
- 2026-04-20
- Provider docs
- deepinfra.com/models
Back of the envelope
≈
Also listed at
- Nvidiafree in · free out
- OrcaRouter$0.172 in · $1.03 out
- SiliconFlow (China)$0.29 in · $1.74 out
- CrofAI$0.35 in · $1.75 out
- OpenRouter$0.385 in · $2.45 out
- SiliconFlow$0.39 in · $2.34 out
- Kilo Gateway$0.39 in · $2.34 out
- Alibaba (China)$0.43 in · $2.58 out
- Wafer$0.43 in · $2.60 out
- DigitalOcean$0.55 in · $3.50 out
- Hugging Face$0.60 in · $3.60 out
- NovitaAI$0.60 in · $3.60 out
- Cortecs$0.60 in · $3.60 out
- Nebius Token Factory$0.60 in · $3.60 out
- Mixlayer$0.60 in · $3.60 out