context.vn

bbh

4 models evaluated

#ModelProviderTypeScore
1DeepSeek V4 Pro BaseDeepSeekOpendeepseek-v4-pro-base
3Soofi S 30B-A3BSoofi ProjectOpensoofi-s-30b-a3b
4MiniCPM5-1BOpenBMB · Open weight71.9%
5Gemma 4 12BGoogle · Open weight53%