Artificial Analysis

GDPval-AA normalized

An Artificial Analysis normalized score for economically valuable tasks.

In index
Agentic
Best in agentic
Source ↗

Models134

Top result68.3%Claude Opus 5.5

Top-3 spread1.0 pts

Year2026

Result and price

100 of 134 results

Results

134 results

#ModelResult

01Claude Opus 5.568.3%

02Muse Wonder Full67.9%

03Kimi K3.1 Pro67.3%

04Claude Sonnet 5.566.9%

05Muse Wonder63.5%

06Claude Fable 5.162.7%

07Claude Opus 561.1%

08Grok 4.760.7%

09MiMo-V2.6-Pro59.2%

10Muse Spark 1.359.0%

11Qwen3.8 Max Preview58.3%

12GLM-5.357.5%

13Qwen3.8-Flash-Next56.4%

14Gemini 4 Argon56.2%

15Claude Haiku 5.556.0%

16Grok 4.655.9%

16Ling 3.1 Flash55.9%

18Claude Fable 555.5%

19GPT-5.6 Sol55.4%

19MiMo-V2.6-Flash55.4%

21DeepSeek V4.1 Flash 101755.0%

22GPT-6.1 Sol54.8%

23DeepSeek V4 Pro 081354.5%

24Step 5 Preview54.3%

25GPT-6 Astra53.7%

26Kimi K351.6%

27GPT-6 Sol50.4%

28Muse Spark 1.248.6%

29Claude Sonnet 548.2%

30GPT-5.6 Luna48.0%

31MiniMax M3.1 Flash47.9%

32Claude Opus 4.847.7%

33GPT-5.6 Terra47.6%

34Gemini 3.8 Flash46.8%

35DeepSeek V4 Flash 073146.7%

36GPT-6 Luna46.6%

37Mistral Large 446.1%

38Qwen3.8-27B45.9%

39Muse Spark Open45.1%

40Gemini 3.7 Flash44.6%

41Grok 4.544.2%

42GLM-5.243.6%

43Claude Opus 4.7 (Adaptive)42.6%

44GPT-5.542.5%

45Gemini 3.5 Flash42.2%

46Gemini 3.6 Flash39.2%

47GPT-5.437.3%

48MiniMax M337.1%

49Hy3 Preview35.8%

50Muse Spark 1.135.6%

51Apodex 1.134.4%

51Apodex 1.1 Mini34.4%

53Quasar 438B34.1%

54Beam 1.0 501B-A23B33.9%

55Nemotron 3.5 Super VL 120B A12B33.1%

55Nemotron 3.5 Super VL 120B A12B NVFP433.1%

55Nemotron 3 Ultra33.1%

58Ling 3.0 Flash VL33.0%

59Qwen3.7 Max31.4%

60Inkling-Small31.2%

61MiMo-V2.5-Pro31.0%

62GLM-5.130.8%

63Solar Pro 430.3%

64Grok 4.329.2%

65Inkling28.6%

66Hy328.2%

67Kimi K2.626.9%

67Kimi K2.7 Code26.9%

69Step 3.7 Flash25.8%

70GLM-4.725.7%

71GPT-5.4 mini25.6%

72MiniMax M2.725.5%

73Muse Spark24.9%

74Qwen3.6 Plus24.5%

75Qwen3.6-27B24.3%

76Gemini 3.5 Flash-Lite24.2%

77Nemotron 3.5 Bolt 74B-A9B24.0%

78A.X K223.0%

79GPT-5.4 nano22.4%

79Ling 3.0 Flash22.4%

81GPT-5 (high)20.9%

82Qwen3.6-35B-A3B19.6%

83Kimi K2.5 (Reasoning)17.0%

83Kimi K2.517.0%

85GPT-5.116.3%

86Qwen3.5-122B-A10B15.7%

87K-EXAONE 2.015.4%

88Gemini 3.1 Pro14.6%

89Muse Glimmer 30B14.2%

90Qwen3.7 Plus13.3%

91Mistral Medium 3.5 128B13.0%

92MiniCPM5-2B10.4%

93Nemotron 3.5 Lightning 30B A3B NVFP46.9%

94MiMo-V2-Flash6.4%

95Gemma 4 31B5.9%

96GPT-OSS 120B5.4%

97Granite 4.2 30B3.7%

98Gemma 4 26B A4B3.2%

99Ling 3.0 Tiny3.0%

100Command A+0.7%

101Gemini 2.5 Pro0.0%

101Mercury 2.50.0%

101Trinity-Large-Thinking0.0%

101K-Exaone0.0%

101Trinity-Large-Preview0.0%

101GPT-OSS 20B0.0%

101Solar Pro 30.0%

101Gemma 4 12B0.0%

101Nemotron 3 Nano Omni 30B A3B0.0%

101North Mini Code0.0%

101Mistral Small 40.0%

101Granite 4.2 8B0.0%

101DeepSeek V3 03240.0%

101Ling 2.6 Flash0.0%

101GPT-4.1 mini0.0%

101Llama 4 Maverick0.0%

101Mistral Large 30.0%

101Celeris-10.0%

101Granite 4.2 3B0.0%

101Nemotron 3 Nano 30B0.0%

101DeepSeek V30.0%

101LFM2.5-2.6B0.0%

101GPT-4o0.0%

101Llama 3.3 70B0.0%

101Gemma 4 E4B0.0%

101Llama 4 Scout0.0%

101GPT-4.1 nano0.0%

101GPT-4o mini0.0%

101Gemma 3 27B0.0%

101Mistral Small 3.10.0%

101Claude 3.5 Haiku0.0%

101Gemma 4 E2B0.0%

—Mistral Small 4 (Reasoning)0.0%

101Ultravox v0.6 Llama 3.3 70B0.0%