context.vn
Menu

mmmu Pro

42 models evaluated

#ModelProviderTypeScore
1Gemini 3.1 ProGoogleProprietary83.9%
2Gemini 3.5 FlashGoogleProprietary83.6%
3GPT-5.6 SolOpenAIProprietary83%
4Qwen3.8 MaxAlibaba · Open weight82.3%
5Kimi K3Moonshot AI · Closed81.6%
6Seed 2.1 ProByteDance · Closed81.6%
7GPT-5.5OpenAI · Closed81.2%
8GPT-5.4OpenAI · Closed81.2%
9Gemini 3 ProGoogle · Closed81%
10GPT-5.6 TerraOpenAI · Closed80.7%
11Muse SparkMeta · Closed80.4%
12Seed 2.1 TurboByteDance · Closed80.1%
13GPT-5.2OpenAI · Closed79.5%
14Kimi K2.6Moonshot AI · Open weight79.4%
15dots3-note PreviewDots Studio · Open weight79.1%
16Qwen3.7 PlusAlibaba · Closed79%
17Qwen3.5 397BAlibaba · Open weight79%
18Qwen3.6 PlusAlibaba · Closed78.8%
19Kimi K2.5Moonshot AI · Open weight78.5%
20Kimi K2.5 (Reasoning)Moonshot AI · Closed78.5%
21GPT-5.6 LunaOpenAI · Closed78.4%
22MiniMax M3MiniMax · Open weight78.1%
23Grok 4.3xAI · Closed78.1%
24Pareto 26.9Unbiased · Closed78%
25MiMo-V2.5Xiaomi · Closed77.9%
26Claude Opus 4.6Anthropic · Closed77.3%
27Gemma 4 31BGoogle · Open weight76.9%
28GPT-5.4 miniOpenAI · Closed76.6%
29Step 5 PreviewStepFun · Closed76%
30Qwen3.6-27BAlibaba · Open weight75.8%
31Qwen3.6-35B-A3BAlibaba · Open weight75.3%
32Grok 4.20xAI · Closed75.2%
33Inkling-SmallThinking Machines Lab · Open weight74%
34Muse Glimmer 30BMeta · Open weight74%
35Gemma 4 26B A4BGoogle · Open weight73.8%
36InklingThinking Machines Lab · Open weight73.5%
37Interfaze BetaInterfaze · Closed71.1%
38Claude Opus 4.5Anthropic · Closed70.6%
39Gemma 4 12BGoogle · Open weight69.1%
40GPT-5.4 nanoOpenAI · Closed66.1%
41Command A+Cohere · Open weight63%
42LFM2.5-VL-3BLiquidAI · Open weight30.5%