Qwen3-VL-32B-Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video. With 32 billion parameters, it combines deep visual perception with advanced text...
Rank
#119
Up 57% week over week
Tokens
17.7B
2026-08-14
Requests
5.2M
5,216,658
Tool calls
-
Input / M
$0.10
Output / M
$0.42
Cache / M
Free
Context
131K
131,072 tokens