Verified record · from the answer index
Qwen: Qwen3.5 397B A17B
modeltested · static-verifiedscore 20/100signed passport ✓published 248d ago · verified 2026-10-20
The 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...
Declared tools (3)
text+image+video->texttool-callingstructured-outputs
model:openrouter:qwen/qwen3.5-397b-a17b