通义听悟
Great for meeting-heavy teams, students, journalists, and researchers turning speech into knowledge; not for formal archival use without human proofreading. Verification level not recorded ·
qwen/qwen3.5-397b-a17b-20260216
Rank #111
qwen
qwen/qwen3.5-397b-a17b-20260216
Usage data date 2026-09-12 · updated daily
Usage and specs
Daily rank#111
Daily tokens41.7B
Daily requests2.7M
Daily change
+52.3%
Released
What is qwen/qwen3.5-397b-a17b-20260216
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...
qwen/qwen3.5-397b-a17b-20260216 capabilities
Input: Text、Image、Video Output: Text
qwen/qwen3.5-397b-a17b-20260216 API pricing
Input · per 1M tokens$0.55
Output · per 1M tokens$3.50
USD list-price snapshot, not independently verified; confirm current pricing and terms on the provider site before purchase.
Tools built on this model
通义听悟
Great for meeting-heavy teams, students, journalists, and researchers turning speech into knowledge; not for formal archival use without human proofreading. Verification level not recorded ·