Qwen3.5 397B A17B
AvailableThe Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers...
- Context
- 262.1K tokens
- Max output
- 65.5K tokens
- Input / 1M tokens
- $0.39per 1M tokens
- Output / 1M tokens
- $2.34per 1M tokens
- Listed since
- Feb 16, 2026
Prices and model data from OpenRouter
Where it's availableProviders · 1
| Provider | Input / 1M tokens | Output / 1M tokens | Status |
|---|---|---|---|
| OpenRouter | $0.39 | $2.34 | Available |
Published benchmark results
From independent leaderboards, each credited to its publisher.