Qwen3.8 27B is an open-weight dense vision-language model from Qwen.
Cost rate
0.4x
Context
262K
Released
Aug 14, 2026
Input
TextImageVideo
Output
Text
Support
ReasoningTool callingStructured outputs
Best at
The categories where this model ranks highest.
Image understanding
#64 路 top 38%
Mathematical
#78 路 top 19%
Software & IT Services
#88 路 top 21%
Life, Physical, & Social Science
#88 路 top 21%