Artificial Analysis

Artificial Analysis MMMU-Pro

An Artificial Analysis MMMU-Pro score.

In index
Multimodal
Best in multimodal
Source ↗

Models114

Top result87.7%Claude Opus 5.5

Top-3 spread1.7 pts

Year2026

Result and price

86 of 114 results

Results

114 results

#ModelResult

01Claude Opus 5.587.7%

02GPT-6 Astra86.9%

03GPT-6.1 Sol86.0%

04Gemini 3.8 Flash85.6%

05Gemini 3.7 Flash85.5%

06GPT-6.1 Terra84.7%

06Claude Opus 584.7%

08Gemini 3.5 Flash84.3%

09Kimi K3.1 Pro83.9%

10GPT-5.6 Sol83.4%

11Gemini 3.6 Flash83.2%

12GPT-6 Sol82.9%

13Qwen3.8 Max Preview82.8%

14Gemini 3.1 Pro82.4%

15Inkling 2 Small81.6%

16Muse Spark Open80.9%

17GPT-5.6 Terra80.7%

18Kimi K380.5%

18Muse Spark80.5%

18Qwen3.7 Plus80.5%

21Grok 4.580.4%

22Gemini 3 Pro80.2%

23GPT-5.579.9%

24Qwen3.8-Flash-Next79.8%

25GPT-6 Luna79.7%

26Kimi K2.679.4%

27Apodex 1.179.2%

27Apodex 1.1 Mini79.2%

29MiniMax M3.1 Flash79.1%

30Gemini 3.5 Flash-Lite79.0%

30Ling 3.0 Flash VL79.0%

32Claude Opus 4.7 (Adaptive)78.8%

33GPT-5.6 Luna78.6%

33MiniMax M378.6%

33Gemini 3 Flash78.6%

36GPT-5.3 Codex78.5%

37GPT-5.478.4%

38Grok 4.378.1%

39Qwen3.6 Plus78.0%

40Claude Sonnet 577.3%

40Qwen3.5 397B (Reasoning)77.3%

40Qwen3.5 397B77.3%

43DeepSeek V4.1 Flash 101777.0%

44Step 5 Preview76.4%

44Claude Opus 4.776.4%

46GPT-5.2-Codex76.3%

46Qwen3.8-27B76.3%

48GPT-5.175.5%

49Claude Opus 4.6 (Adaptive)75.4%

49Kimi K2.5 (Reasoning)75.4%

49Kimi K2.575.4%

52Step 3.7 Flash75.3%

53Qwen3.5-122B-A10B75.0%

53Qwen3.5-27B75.0%

53Qwen3.6-35B-A3B75.0%

56Gemini 2.5 Pro74.9%

57Qwen3.6-27B74.6%

58Muse Glimmer 30B74.3%

58GPT-5 (medium)74.3%

60GPT-5 (high)74.2%

61Claude Opus 4.5 Thinking74.0%

61Inkling-Small74.0%

63Inkling73.5%

64Gemma 4 31B73.4%

65GPT-5.4 mini73.3%

66MiMo-V2.6-Flash73.1%

67GLM-5V-Turbo72.8%

68Qwen3.5-35B-A3B72.7%

69Claude Opus 4.672.5%

69GPT-5.1-Codex-Max72.5%

69GPT-5.1-Codex72.5%

72Claude Opus 4.571.2%

73Claude Sonnet 4.670.6%

74o370.1%

75MiMo-V2-Omni69.9%

76Gemma 4 12B69.7%

77o4-mini69.2%

77Gemma 4 26B A4B69.2%

79Grok 468.8%

80Claude 4.1 Opus Thinking67.9%

81Gemini 2.5 Flash65.5%

82GPT-5.4 nano65.4%

82Qwen3.5-4B65.4%

84Mistral Medium 3.5 128B64.9%

85Grok 4.1 Fast (Reasoning)63.3%

86Command A+63.2%

87Claude 4 Sonnet62.4%

88Llama 4 Maverick62.1%

89Grok 4 Fast (Reasoning)61.8%

90GPT-4.161.2%

91Qwen3-Omni-30B-A3B-Thinking60.2%

92Claude 3.7 Sonnet60.1%

93GPT-4.1 mini58.7%

94Mistral Small 456.8%

—Mistral Small 4 (Reasoning)56.8%

95Mistral Large 355.7%

96Qwen3-Omni-30B-A3B-Instruct55.5%

97Gemini 1.5 Pro55.0%

98Gemini 2.5 Flash-Lite54.0%

99Nemotron 3 Nano Omni 30B A3B53.2%

100Mistral Medium 353.0%

101Llama 4 Scout52.9%

102Gemma 4 E4B51.4%

103Grok 4.1 Fast48.4%

103Gemini 1.5 Flash48.4%

105Gemma 3 27B48.0%

106Claude 3.5 Haiku45.6%

107Gemma 4 E2B44.6%

108Nova Pro44.3%

109GPT-4o mini41.5%

110GPT-4.1 nano40.1%

111Claude 3 Haiku30.8%

112LFM2.5-VL-1.6B-Extract26.5%

113Phi-4 Multimodal Instruct14.5%