The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the...
Modalities
TextImageVideo
In / out price
₩114 / ₩456
1M
Context
100万
リリース日
2026/02/26
アクティビティデータはありません。