Type to search · Enter for full results

8.5 Global

mistral-small3.2:latest

Judge: gemma4:31b · 152/160 tests · 1 h 1 min · 14.7 tok/s

24.0B · Q4_K_M · 14.1 GB · 131K ctx

VisionTools

Category breakdown

surprise 10.0
agentic 9.8
reasoning 9.6
vision 9.2
long-context 9.1
frontend 8.8
instruction 8.8
code 8.6
multilingual 8.5
roleplay 8.5
math 8.0
organization 7.6
writing 7.3
safety 7.1