The Qwen3.5 122B-A10B 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. In terms of...
Rank
#129
Up 17% week over week
Tokens
15.5B
2026-08-14
Requests
2.3M
2,333,503
Tool calls
-
Input / M
$0.29
Output / M
$2.40
Cache / M
Free
Context
262K
262,144 tokens