artificialanalysis.ai · benchmark source

Artificial Analysis

Independent evaluations across composite intelligence, science reasoning, and agentic terminal tasks. Multiple boards live under one publisher.

Open artificialanalysis.ai3 boards on this publisher

Expert reasoning

GPQA Diamond

A difficult question set intended to test graduate-level scientific reasoning.

Why it matters

It helps separate broad reasoning ability from surface fluency on demanding knowledge questions.

Use it when

Your work includes difficult science reasoning and you can tolerate a narrow benchmark lens.

The limitation

Scientific question answering is not the same as coding, writing, retrieval, or safe production use.

Leaderboard snapshot

GPQA Diamond results in a readable view

Use the tabs for the primary score and other published fields from this board. The source link remains authoritative for the live table.

Full Artificial Analysis GPQA Diamond table mirrored.

Local leader

Grok 4.6 (high)

94.9%

Rows shown

565

Accuracy

Snapshot date

2026-08-14

1 day old · not a live API feed

Score profile

Accuracy by published row

Higher is better in this view

Grok 4.6 (high)

94.9%

Gemini 3.7 Flash (high)

94.5%

Gemini 3.1 Pro Preview

94.1%

GPT-5.6 Sol (max)

94.1%

Claude Opus 5 (Adaptive Reasoning, High Effort)

93.7%

Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)

93.7%

GPT-5.5 (xhigh)

93.5%

Kimi K3 (max)

93.5%

Claude Opus 5 (Adaptive Reasoning, Max Effort)

93.2%

GPT-5.5 (high)

93.2%

GPT-5.6 Sol (xhigh)

93.1%

Grok 4.5 (high)

93.1%

MiniMax-M3

92.9%

DeepSeek V4 Pro 0813 (max)

92.8%

Gemini 3.6 Flash (high)

92.8%

GPT-5.6 Sol (high)

92.8%

Qwen3.8 Max

92.7%

Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)

92.6%

GPT-5.5 (medium)

92.6%

GPT-5.6 Sol (medium)

92.6%

GPT-5.6 Terra (max)

92.5%

Qwen3.7 Max

92.3%

Gemini 3.5 Flash (high)

92.2%

Gemini 3.5 Flash (medium)

92.1%

Claude Opus 4.8 (Adaptive Reasoning, Max Effort)

92%

GPT-5.4 (xhigh)

92%

Claude Opus 5 (Adaptive Reasoning, Medium Effort)

91.9%

GPT-5.3 Codex (xhigh)

91.5%

Claude Opus 4.7 (Adaptive Reasoning, Max Effort)

91.4%

Claude Sonnet 5 (Adaptive Reasoning, Max Effort)

91.1%

GPT-5.6 Luna (max)

91.1%

Grok 4.20 0309 v2 (Reasoning)

91.1%

Kimi K2.6

91.1%

GPT-5.5 (low)

91%

DeepSeek V4 Flash 0731 (Reasoning, Max Effort)

90.8%

Gemini 3 Pro Preview (high)

90.8%

GPT-5.6 Terra (xhigh)

90.8%

DeepSeek V4 Pro (Reasoning, High Effort)

90.5%

Muse Spark 1.2 (xhigh)

90.4%

GPT-5.2 (xhigh)

90.3%

Grok 4.3 (high)

90.1%

Qwen3.7 Plus

90%

GPT-5.2 Codex (xhigh)

89.9%

Gemini 3 Flash Preview (Reasoning)

89.8%

GPT-5.6 Sol (low)

89.8%

Muse Spark 1.1 (xhigh)

89.8%

Hy3

89.7%

Claude Opus 4.6 (Adaptive Reasoning, Max Effort)

89.6%

GPT-5.6 Terra (high)

89.6%

Kimi K2.7 Code

89.6%

GLM-5.2 (max)

89.5%

GPT-5.6 Luna (xhigh)

89.5%

Grok Build 0.1 0616

89.5%

Inkling Small

89.5%

DeepSeek V4 Flash (Reasoning, Max Effort)

89.4%

Qwen3.5 397B A17B (Reasoning)

89.3%

GPT-5.6 Luna (high)

89.2%

Nex-N2-Pro

89.2%

Grok 4.3 (medium)

89%

Claude Opus 5 (Adaptive Reasoning, Low Effort)

88.9%

Qwen3.6 Max Preview

88.8%

Gemini 3 Pro Preview (low)

88.7%

Claude Opus 4.7 (Non-reasoning, High Effort)

88.5%

Grok 4.20 0309 (Reasoning)

88.5%

Muse Spark

88.4%

Qwen3.6 Plus

88.2%

Kimi K2.5 (Reasoning)

87.9%

Grok 4

87.7%

Agnes 2.5 Pro Alpha

87.6%

Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

87.5%

GPT-5.4 mini (xhigh)

87.5%

MiniMax-M2.7

87.4%

GPT-5.1 (high)

87.3%

GPT-5.6 Terra (medium)

87.2%

Inkling (xhigh)

87.2%

DeepSeek V3.2 Speciale

87.1%

GPT-5.4 (low)

87.1%

MiMo-V2-Pro

87%

Motif 3 (Beta)

86.9%

GLM-5.1 (Reasoning)

86.8%

DeepSeek V4 Flash (Reasoning, High Effort)

86.7%

Hy3-preview (Reasoning)

86.7%

Nemotron 3 Ultra 550B A55B (Reasoning)

86.7%

Claude Opus 4.5 (Reasoning)

86.6%

MiMo-V2.5-Pro

86.6%

GPT-5.2 (medium)

86.4%

Qwen3 Max Thinking

86.1%

Qwen3.5 397B A17B (Non-reasoning)

86.1%

GPT-5.1 Codex (high)

86%

GLM-4.7 (Reasoning)

85.9%

GPT-5.6 Luna (medium)

85.9%

Qwen3.5 27B (Reasoning)

85.8%

Gemma 4 31B (Reasoning)

85.7%

Qwen3.5 122B A10B (Reasoning)

85.7%

Ring-2.6-1T

85.7%

KAT Coder Pro V2

85.5%

Ling 3.0 Flash

85.5%

MiMo-V2-Omni-0327

85.5%

GPT-5 (high)

85.4%

Grok 4.1 Fast (Reasoning)

85.3%

MiMo-V2.5

84.9%

Nanbeige4.1-3B

84.9%

MiniMax-M2.5

84.8%

GLM-5-Turbo

84.7%

Grok 4 Fast (Reasoning)

84.7%

GPT-5.5 Instant (May 2026)

84.6%

MiMo-V2-Flash (Reasoning)

84.6%

JT-4.1 Flash 236B A21B

84.5%

o3-pro

84.5%

Qwen3.5 35B A3B (Reasoning)

84.5%

Gemini 2.5 Pro

84.4%

GPT-5.6 Terra (low)

84.3%

Grok 4.3 (low)

84.3%

GPT-5 (medium)

84.2%

Kimi K3 (low)

84.2%

Qwen3.5 27B (Non-reasoning)

84.2%

Qwen3.6 27B (Reasoning)

84.2%

Qwen3.6 35B A3B (Reasoning)

84.1%

Claude Opus 4.6 (Non-reasoning, High Effort)

84%

DeepSeek V3.2 (Reasoning)

84%

GLM-5.1 (Non-reasoning)

83.9%

Gemini 3.5 Flash-Lite

83.8%

Kimi K2 Thinking

83.8%

GPT-5 Codex (high)

83.7%

Gemini 2.5 Pro Preview (Mar' 25)

83.6%

GPT-5.6 Luna (low)

83.5%

MiMo-V2-Flash (Feb 2026)

83.5%

Muse Glimmer (high)

83.5%

Claude 4.5 Sonnet (Reasoning)

83.4%

Step 3.5 Flash

83.1%

MiniMax-M2.1

83%

JT-35B-Flash

82.9%

Qwen3.6 27B (Non-reasoning)

82.9%

Gemini 3.5 Flash (minimal)

82.8%

GPT-5 mini (high)

82.8%

MiMo-V2-Omni

82.8%

82.7%

Qwen3.5 122B A10B (Non-reasoning)

82.7%

Qwen3.5 Omni Plus

82.6%

Step 3.5 Flash 2603

82.6%

GPT-5.4 mini (medium)

82.3%

GPT-5.5 Instant (June 2026)

82.3%

Gemini 2.5 Pro Preview (May' 25)

82.2%

Gemini 3.1 Flash-Lite

82.2%

GLM-5 (Reasoning)

82%

Qwen3.5 35B A3B (Non-reasoning)

81.9%

GPT-5.4 nano (xhigh)

81.7%

Qwen3.6 35B A3B (Non-reasoning)

81.7%

DeepSeek R1 0528 (May '25)

81.3%

GPT-5.1 Codex mini (high)

81.3%

Gemini 3 Flash Preview (Non-reasoning)

81.2%

ERNIE 4.5 300B A47B

81.1%

Nova 2.0 Lite (high)

81.1%

Claude Opus 4.5 (Non-reasoning)

81%

Claude 4.1 Opus (Reasoning)

80.9%

GLM 5V Turbo (Reasoning)

80.9%

Step 3.7 Flash

80.9%

GPT-5 (low)

80.8%

Qwen3.5 9B (Reasoning)

80.6%

GPT-5 mini (medium)

80.3%

Claude Sonnet 5 (Non-reasoning, High Effort)

80%

Nemotron 3 Super 120B A12B (Reasoning)

80%

Claude Sonnet 4.6 (Non-reasoning, High Effort)

79.9%

Claude Sonnet 4.6 (Non-reasoning, Low Effort)

79.7%

DeepSeek V3.2 Exp (Reasoning)

79.7%

Claude 4 Opus (Reasoning)

79.6%

EXAONE 4.5 33B

79.4%

Gemini 2.5 Flash Preview (Sep '25) (Reasoning)

79.3%

DeepSeek V3.1 Terminus (Reasoning)

79.2%

Gemma 4 26B A4B (Reasoning)

79.2%

Grok 3 mini Reasoning (high)

79.1%

Gemini 2.5 Flash (Reasoning)

79%

GPT-5.6 Sol (Non-reasoning)

79%

Qwen3 235B A22B 2507 (Reasoning)

79%

Kimi K2.5 (Non-reasoning)

78.9%

Kimi K2.6 (Non-reasoning)

78.8%

Qwen3.5 9B (Non-reasoning)

78.6%

Grok 4.20 0309 (Non-reasoning)

78.5%

Nova 2.0 Pro Preview (medium)

78.5%

o4-mini (high)

78.4%

K-EXAONE (Reasoning)

78.3%

GLM-4.5 (Reasoning)

78.2%

gpt-oss-120b (high)

78.2%

GLM-4.6 (Reasoning)

78%

LongCat 2.0

78%

DeepSeek V3.1 (Reasoning)

77.9%

Claude 4 Sonnet (Reasoning)

77.7%

ERNIE 5.0 Thinking Preview

77.7%

MiniMax-M2

77.7%

Grok 4.20 0309 v2 (Non-reasoning)

77.6%

Qwen3 Max Thinking (Preview)

77.6%

Ring-1T

77.4%

o3-mini (high)

77.3%

Claude 3.7 Sonnet (Reasoning)

77.2%

Qwen3 VL 235B A22B (Reasoning)

77.2%

Qwen3.5 4B (Reasoning)

77.1%

Mercury 2

77%

Mistral Small 4 (Reasoning)

76.9%

Cogito v2.1 (Reasoning)

76.8%

GPT-5.5 (Non-reasoning)

76.8%

Nova 2.0 Lite (medium)

76.8%

Kimi K2 0905

76.7%

Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)

76.6%

Kimi K2

76.6%

Doubao Seed Code

76.4%

KAT-Coder-Pro V1

76.4%

Qwen3 Max

76.4%

Qwen3 Max (Preview)

76.4%

Gemma 4 31B (Non-reasoning)

76.3%

MiMo-V2.5-Pro (Non-reasoning)

76.2%

Command A+

76.1%

GPT-5.4 nano (medium)

76.1%

INTELLECT-3

76.1%

Nova 2.0 Omni (medium)

76%

Qwen3 Next 80B A3B (Reasoning)

75.9%

Nemotron Cascade 2 30B A3B

75.8%

North Mini Code

75.7%

NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)

75.7%

G9v3-39A5B

75.6%

Gemma 4 12B (Reasoning)

75.3%

Qwen3 235B A22B 2507 Instruct

75.3%

Ling-2.6-1T

75.2%

Trinity Large Thinking

75.2%

DeepSeek V3.1 Terminus (Non-reasoning)

75.1%

DeepSeek V3.2 (Non-reasoning)

75.1%

Nova 2.0 Pro Preview (low)

75.1%

GPT-5.4 (Non-reasoning)

74.8%

Llama Nemotron Super 49B v1.5 (Reasoning)

74.8%

Mistral Medium 3.5

74.8%

o3-mini

74.8%

74.7%

GPT-5.6 Terra (Non-reasoning)

74.6%

Nemotron 3.5 Lightning

74.3%

Qwen3.5 Omni Flash

74.2%

EXAONE 4.0 32B (Reasoning)

73.9%

Magistral Medium 1.2

73.9%

DeepSeek V3.2 Exp (Non-reasoning)

73.8%

Qwen3 Next 80B A3B Instruct

73.8%

Sarvam 105B (high)

73.8%

Qwen3 Coder Next

73.7%

DeepSeek V3.1 (Non-reasoning)

73.5%

Ling 3.0 Tiny

73.4%

Apriel-v1.6-15B-Thinker

73.3%

GLM-4.5-Air

73.3%

HyperNova 60B 2605

73.3%

Qwen3 VL 32B (Reasoning)

73.3%

Hy3-preview (Non-reasoning)

73.2%

Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)

72.8%

Claude 4.5 Sonnet (Non-reasoning)

72.7%

Grok Code Fast 1

72.7%

Hermes 4 - Llama-3.1 405B (Reasoning)

72.7%

Qwen3 Omni 30B A3B (Reasoning)

72.6%

Seed-OSS-36B-Instruct

72.6%

Ring-flash-2.0

72.5%

Solar Pro 3

72.4%

Mi:dm K 2.5 Pro Preview

72.2%

Qwen3 VL 30B A3B (Reasoning)

72%

GLM-4.6V (Reasoning)

71.9%

Ling-1T

71.9%

DeepSeek V4 Pro (Non-reasoning)

71.7%

DeepSeek V4 Flash (Non-reasoning)

71.6%

Gemma 4 26B A4B (Non-reasoning)

71.4%

Apriel-v1.5-15B-Thinker

71.3%

K2 Think V2

71.3%

GPT-5.2 (Non-reasoning)

71.2%

Qwen3 VL 235B A22B Instruct

71.2%

Qwen3.5 4B (Non-reasoning)

71.2%

Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)

70.9%

DeepSeek R1 (Jan '25)

70.8%

Qwen3 30B A3B 2507 (Reasoning)

70.7%

Claude 4 Opus (Non-reasoning)

70.1%

Gemini 2.0 Flash Thinking Experimental (Jan '25)

70.1%

Mi:dm K 2.5 Pro

70.1%

Qwen3 235B A22B (Reasoning)

70%

Hermes 4 - Llama-3.1 70B (Reasoning)

69.9%

Nova 2.0 Omni (low)

69.9%

Gemini 2.5 Flash Preview (Reasoning)

69.8%

Nova 2.0 Lite (low)

69.8%

MiniMax M1 80k

69.7%

K-EXAONE (Non-reasoning)

69.5%

Motif-2-12.7B-Reasoning

69.5%

Qwen3 VL 30B A3B Instruct

69.5%

Grok 3

69.3%

Step3 VL 10B

69%

gpt-oss-20b (high)

68.8%

GPT-5 mini (minimal)

68.7%

Solar Pro 2 (Reasoning)

68.7%

GLM-5.2 (Non-reasoning)

68.6%

GPT-5 (ChatGPT)

68.6%

GLM-4.5V (Reasoning)

68.4%

Claude 4 Sonnet (Non-reasoning)

68.3%

Gemini 2.5 Flash (Non-reasoning)

68.3%

MiniMax M1 40k

68.2%

K2-V2 (high)

68.1%

Mistral Large 3

68%

Magistral Medium 1

67.9%

GPT-5 nano (high)

67.6%

JT-MINI

67.6%

GPT-5 (minimal)

67.3%

Claude 4.5 Haiku (Reasoning)

67.2%

gpt-oss-120b (low)

67.2%

Llama 4 Maverick

67.1%

Qwen3 VL 32B Instruct

67.1%

GPT-5 nano (medium)

67%

DiffusionGemma 26B A4B

66.9%

Qwen3 32B (Reasoning)

66.8%

Qwen3 4B 2507 (Reasoning)

66.7%

GLM-5 (Non-reasoning)

66.6%

GPT-4.1

66.6%

GLM-4.7 (Non-reasoning)

66.4%

GPT-4.1 mini

66.4%

Magistral Small 1.2

66.3%

Falcon-H1R-7B

66.1%

Gemma 4 12B (Non-reasoning)

66.1%

Qwen3 30B A3B 2507 Instruct

65.9%

Grok 4.3 (Non-reasoning)

65.8%

Ling-flash-2.0

65.7%

Solar Open 100B (Reasoning)

65.7%

Claude 3.7 Sonnet (Non-reasoning)

65.6%

MiMo-V2-Flash (Non-reasoning)

65.6%

DeepSeek V3 0324

65.5%

GPT-4o (March 2025, chatgpt-4o-latest)

65.5%

Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)

65.1%

Claude 4.5 Haiku (Non-reasoning)

64.6%

GPT-5.6 Luna (Non-reasoning)

64.5%

GPT-5.1 (Non-reasoning)

64.3%

Llama 3.3 Nemotron Super 49B v1 (Reasoning)

64.3%

Magistral Small 1

64.1%

Grok 4.1 Fast (Non-reasoning)

63.7%

Gemini 2.0 Flash (experimental)

63.6%

LongCat Flash Lite

63.6%

Nova 2.0 Pro Preview (Non-reasoning)

63.6%

Sarvam 30B (high)

63.3%

GLM-4.6 (Non-reasoning)

63.2%

Celeris-1

63.1%

EXAONE 4.0 32B (Non-reasoning)

62.8%

Gemini 2.5 Flash-Lite (Reasoning)

62.5%

Gemini 2.0 Flash (Feb '25)

62.3%

Sonar Reasoning

62.3%

Gemini 2.0 Pro Experimental (Feb '25)

62.2%

Qwen3 Omni 30B A3B Instruct

62%

Qwen3 Coder 480B A35B Instruct

61.8%

Qwen3 30B A3B (Reasoning)

61.6%

DeepSeek R1 Distill Qwen 32B

61.5%

HyperCLOVA X SEED Think (32B)

61.5%

Qwen3 235B A22B (Non-reasoning)

61.3%

DeepSeek R1 0528 Qwen3 8B

61.2%

gpt-oss-20b (low)

61.1%

Olmo 3 32B Think

61%

GPT-5.4 mini (Non-Reasoning)

60.6%

Grok 4 Fast (Non-reasoning)

60.6%

Qwen3 14B (Reasoning)

60.4%

Nova 2.0 Lite (Non-reasoning)

60.3%

o1-mini

60.3%

Tri-21B-Think

60.1%

Claude 3.5 Sonnet (Oct '24)

59.9%

K2-V2 (medium)

59.8%

Devstral 2

59.4%

Gemini 2.5 Flash Preview (Non-reasoning)

59.4%

Ling 2.6 Flash

59.3%

QwQ 32B

59.3%

Olmo 3.1 32B Think

59.1%

Gemini 1.5 Pro (Sep '24)

58.9%

Qwen3 8B (Reasoning)

58.9%

Mistral Medium 3.1

58.8%

Llama 4 Scout

58.7%

Qwen2.5 Max

58.7%

GLM-4.7-Flash (Reasoning)

58.1%

Qwen3 VL 8B (Reasoning)

57.9%

Mistral Medium 3

57.8%

Solar Pro 2 (Preview) (Reasoning)

57.8%

Sonar Pro

57.8%

Gemma 4 E4B (Reasoning)

57.6%

Phi-4

57.5%

GLM-4.5V (Non-reasoning)

57.3%

Ministral 3 14B

57.2%

NVIDIA Nemotron Nano 12B v2 VL (Reasoning)

57.2%

Mistral Small 4 (Non-reasoning)

57.1%

NVIDIA Nemotron Nano 9B V2 (Reasoning)

57%

Nova Premier

56.9%

GLM-4.6V (Non-reasoning)

56.6%

Ling-mini-2.0

56.2%

Solar Pro 2 (Non-reasoning)

56.1%

Claude 3.5 Sonnet (June '24)

56%

GPT-5.4 nano (Non-Reasoning)

55.8%

DeepSeek V3 (Dec '24)

55.7%

NVIDIA Nemotron Nano 9B V2 (Non-reasoning)

55.7%

QwQ 32B-Preview

55.7%

Nova 2.0 Omni (Non-reasoning)

55.5%

Gemma 4 E4B (Non-reasoning)

54.9%

Solar Pro 2 (Preview) (Non-reasoning)

54.4%

GPT-4o (Nov '24)

54.3%

Gemini 2.0 Flash-Lite (Preview)

54.2%

K2-V2 (low)

54.1%

Olmo 3.1 32B Instruct

53.9%

Tri-21B-think Preview

53.8%

Hermes 4 - Llama-3.1 405B (Non-reasoning)

53.6%

Gemini 2.0 Flash-Lite (Feb '25)

53.5%

Qwen3 32B (Non-reasoning)

53.5%

Devstral Small 2

53.2%

Reka Flash 3

52.9%

Command A

52.7%

GPT-4o (May '24)

52.6%

Qwen3 4B (Reasoning)

52.2%

GPT-4o (Aug '24)

52.1%

Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)

51.7%

Qwen3 4B 2507 Instruct

51.7%

Llama 3.1 Tulu3 405B

51.6%

Olmo 3 7B Think

51.6%

Qwen3 Coder 30B A3B Instruct

51.6%

Exaone 4.0 1.2B (Reasoning)

51.5%

Llama 3.1 Instruct 405B

51.5%

Qwen3 30B A3B (Non-reasoning)

51.5%

LFM2.5-8B-A1B

51.3%

NVIDIA Nemotron 3 Nano 4B

51.3%

GPT-4.1 nano

51.2%

GPT-4o (ChatGPT)

51.1%

Grok 2 (Dec '24)

51%

Mistral Small 3.2

50.5%

Pixtral Large

50.5%

Nova Pro

49.9%

Llama 3.3 Instruct 70B

49.8%

Qwen3 VL 4B (Reasoning)

49.4%

Devstral Medium

49.2%

Hermes 4 - Llama-3.1 70B (Non-reasoning)

49.1%

Qwen2.5 Instruct 72B

49.1%

Claude 3 Opus

48.9%

Mistral Large 2 (Nov '24)

48.6%

DeepSeek R1 Distill Qwen 14B

48.4%

Granite 4.1 30B

48.1%

Llama Nemotron Super 49B v1.5 (Non-reasoning)

48.1%

Gemini 2.5 Flash-Lite (Non-reasoning)

47.4%

LFM2 24B A2B

47.4%

Mistral Large 2 (Jul '24)

47.2%

Grok Beta

47.1%

Ministral 3 8B

47.1%

Sonar

47.1%

Qwen3 14B (Non-reasoning)

47%

Nemotron 3 Nano Omni 30B A3B Reasoning

46.9%

Qwen2.5 Instruct 32B

46.6%

Llama 3.1 Nemotron Instruct 70B

46.5%

Gemini 1.5 Flash (Sep '24)

46.3%

Mistral Small 3

46.2%

Qwen3.5 2B (Reasoning)

45.6%

Mistral Small 3.1

45.4%

GLM-4.7-Flash (Non-reasoning)

45.2%

Qwen3 8B (Non-reasoning)

45.2%

NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)

43.9%

G9v3-3B

43.8%

Qwen3.5 2B (Non-reasoning)

43.8%

Devstral Small (May '25)

43.4%

Gemma 4 E2B (Reasoning)

43.3%

Granite 4.1 8B

43.3%

Nova Lite

43.3%

Llama 3.2 Instruct 90B (Vision)

43.2%

Gemma 3 27B Instruct

42.8%

GPT-5 nano (minimal)

42.8%

Jamba 1.5 Large

42.7%

Qwen3 VL 8B Instruct

42.7%

GPT-4o mini

42.6%

Molmo2-8B

42.5%

Exaone 4.0 1.2B (Non-reasoning)

42.4%

Mistral Saba

42.4%

Qwen2.5 Coder Instruct 32B

41.7%

Granite 4.0 H Small

41.6%

Sarvam M (Reasoning)

41.6%

Devstral Small (Jul '25)

41.4%

Kimi Linear 48B A3B Instruct

41.2%

Qwen2.5 Turbo

41%

Llama 3.1 Instruct 70B

40.9%

Claude 3.5 Haiku

40.8%

Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)

40.8%

Gemma 4 E2B (Non-reasoning)

40.5%

DeepSeek R1 Distill Llama 70B

40.2%

Hermes 3 - Llama-3.1 70B

40.1%

Claude 3 Sonnet

40%

Olmo 3 7B Instruct

40%

NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)

39.9%

Qwen3 4B (Non-reasoning)

39.8%

Jamba 1.7 Large

39%

Jamba 1.6 Large

38.7%

DeepHermes 3 - Mistral 24B Preview (Non-reasoning)

38.2%

Mistral Small (Sep '24)

38.1%

Llama 3 Instruct 70B

37.9%

Claude 3 Haiku

37.4%

Gemini 1.5 Pro (May '24)

37.1%

Qwen2 Instruct 72B

37.1%

Qwen3 VL 4B Instruct

37.1%

Gemini 1.5 Flash-8B

35.9%

Ministral 3 3B

35.8%

Nova Micro

35.8%

Qwen3 1.7B (Reasoning)

35.6%

Mistral Large (Feb '24)

35.1%

Gemma 3 12B Instruct

34.9%

GPT-4

34.9%

Mistral Medium

34.9%

Claude 2.0

34.4%

LFM2 8B A1B

34.4%

LFM2.5-1.2B-Thinking

33.9%

Qwen2.5 Coder Instruct 7B

33.9%

Granite 3.3 8B (Non-reasoning)

33.8%

Granite 4.0 Micro

33.6%

Jamba Reasoning 3B

33.3%

Mixtral 8x22B Instruct

33.2%

DBRX Instruct

33.1%

Phi-4 Mini Instruct

33.1%

Claude Instant

33%

OLMo 2 32B

32.8%

LFM 40B

32.7%

Llama 2 Chat 70B

32.7%

LFM2.5-1.2B-Instruct

32.6%

Gemini 1.5 Flash (May '24)

32.4%

Command-R+ (Apr '24)

32.3%

Jamba 1.7 Mini

32.2%

Llama 2 Chat 13B

32.1%

Claude 2.1

31.9%

DeepSeek Coder V2 Lite Instruct

31.9%

Phi-3 Mini Instruct 3.8B

31.9%

Phi-4 Multimodal Instruct

31.5%

Granite 4.1 3B

31.4%

LFM2 2.6B

30.6%

MiniCPM-V 4.6 1.3B

30.5%

Tiny Aya Global

30.5%

DeepSeek R1 Distill Llama 8B

30.2%

Jamba 1.5 Mini

30.2%

Mistral Small (Feb '24)

30.2%

Jamba 1.6 Mini

30%

GPT-3.5 Turbo

29.7%

Gemma 3n E4B Instruct

29.6%

Llama 3 Instruct 8B

29.6%

Mixtral 8x7B Instruct

29.2%

Gemma 3 4B Instruct

29.1%

LFM2.5-VL-1.6B

28.9%

Qwen1.5 Chat 110B

28.9%

OLMo 2 7B

28.8%

Command-R (Mar '24)

28.4%

Qwen3 1.7B (Non-reasoning)

28.3%

Granite 4.0 1B

28.1%

Gemma 3n E4B Instruct Preview (May '25)

27.8%

MiniCPM5-1B (Reasoning)

27.8%

Gemini 1.0 Pro

27.7%

Apertus 70B Instruct

27.2%

DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)

27%

MiniCPM5-1B (Non-reasoning)

26.9%

Granite 4.0 H 1B

26.3%

Granite 4.0 350M

26.1%

Llama 3.1 Instruct 8B

25.9%

Granite 4.0 H 350M

25.7%

Apertus 8B Instruct

25.6%

Llama 3.2 Instruct 3B

25.5%

Molmo 7B-D

24%

Qwen3 0.6B (Reasoning)

23.9%

Gemma 3 1B Instruct

23.7%

Qwen3.5 0.8B (Non-reasoning)

23.6%

Qwen3 0.6B (Non-reasoning)

23.1%

OpenChat 3.5 (1210)

23%

Gemma 3n E2B Instruct

22.9%

LFM2 1.2B

22.8%

Llama 2 Chat 7B

22.7%

Gemma 3 270M

22.4%

Llama 3.2 Instruct 11B (Vision)

22.1%

Llama 3.2 Instruct 1B

19.6%

Mistral 7B Instruct

17.7%

Qwen3.5 0.8B (Reasoning)

11.1%

DeepSeek R1 Distill Qwen 1.5B

9.8%

Benchmark leaderboard snapshot showing rank, score, published details, and notes
Rank	Model or system	Accuracy	Other details	Note
#1	Grok 4.6 (high)SpaceXAI	94.9%	View: Artificial Analysis	—
#2	Gemini 3.7 Flash (high)Google	94.5%	View: Artificial Analysis	—
#3	Gemini 3.1 Pro PreviewGoogle	94.1%	View: Artificial Analysis	—
#3	GPT-5.6 Sol (max)OpenAI	94.1%	View: Artificial Analysis	—
#5	Claude Opus 5 (Adaptive Reasoning, High Effort)Anthropic	93.7%	View: Artificial Analysis	—
#5	Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)Anthropic	93.7%	View: Artificial Analysis	—
#7	GPT-5.5 (xhigh)OpenAI	93.5%	View: Artificial Analysis	—
#7	Kimi K3 (max)Moonshot AI	93.5%	View: Artificial Analysis	—
#9	Claude Opus 5 (Adaptive Reasoning, Max Effort)Anthropic	93.2%	View: Artificial Analysis	—
#9	GPT-5.5 (high)OpenAI	93.2%	View: Artificial Analysis	—
#11	GPT-5.6 Sol (xhigh)OpenAI	93.1%	View: Artificial Analysis	—
#11	Grok 4.5 (high)SpaceXAI	93.1%	View: Artificial Analysis	—
#13	MiniMax-M3MiniMax	92.9%	View: Artificial Analysis	—
#14	DeepSeek V4 Pro 0813 (max)DeepSeek	92.8%	View: Artificial Analysis	—
#14	Gemini 3.6 Flash (high)Google	92.8%	View: Artificial Analysis	—
#14	GPT-5.6 Sol (high)OpenAI	92.8%	View: Artificial Analysis	—
#17	Qwen3.8 MaxQwen	92.7%	View: Artificial Analysis	—
#18	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)Anthropic	92.6%	View: Artificial Analysis	—
#18	GPT-5.5 (medium)OpenAI	92.6%	View: Artificial Analysis	—
#18	GPT-5.6 Sol (medium)OpenAI	92.6%	View: Artificial Analysis	—
#21	GPT-5.6 Terra (max)OpenAI	92.5%	View: Artificial Analysis	—
#22	Qwen3.7 MaxQwen	92.3%	View: Artificial Analysis	—
#23	Gemini 3.5 Flash (high)Google	92.2%	View: Artificial Analysis	—
#24	Gemini 3.5 Flash (medium)Google	92.1%	View: Artificial Analysis	—
#25	Claude Opus 4.8 (Adaptive Reasoning, Max Effort)Anthropic	92%	View: Artificial Analysis	—
#25	GPT-5.4 (xhigh)OpenAI	92%	View: Artificial Analysis	—
#27	Claude Opus 5 (Adaptive Reasoning, Medium Effort)Anthropic	91.9%	View: Artificial Analysis	—
#28	GPT-5.3 Codex (xhigh)OpenAI	91.5%	View: Artificial Analysis	—
#29	Claude Opus 4.7 (Adaptive Reasoning, Max Effort)Anthropic	91.4%	View: Artificial Analysis	—
#30	Claude Sonnet 5 (Adaptive Reasoning, Max Effort)Anthropic	91.1%	View: Artificial Analysis	—
#30	GPT-5.6 Luna (max)OpenAI	91.1%	View: Artificial Analysis	—
#30	Grok 4.20 0309 v2 (Reasoning)SpaceXAI	91.1%	View: Artificial Analysis	—
#30	Kimi K2.6Moonshot AI	91.1%	View: Artificial Analysis	—
#34	GPT-5.5 (low)OpenAI	91%	View: Artificial Analysis	—
#35	DeepSeek V4 Flash 0731 (Reasoning, Max Effort)DeepSeek	90.8%	View: Artificial Analysis	—
#35	Gemini 3 Pro Preview (high)Google	90.8%	View: Artificial Analysis	—
#35	GPT-5.6 Terra (xhigh)OpenAI	90.8%	View: Artificial Analysis	—
#38	DeepSeek V4 Pro (Reasoning, High Effort)DeepSeek	90.5%	View: Artificial Analysis	—
#39	Muse Spark 1.2 (xhigh)Meta	90.4%	View: Artificial Analysis	—
#40	GPT-5.2 (xhigh)OpenAI	90.3%	View: Artificial Analysis	—
#41	Grok 4.3 (high)SpaceXAI	90.1%	View: Artificial Analysis	—
#42	Qwen3.7 PlusQwen	90%	View: Artificial Analysis	—
#43	GPT-5.2 Codex (xhigh)OpenAI	89.9%	View: Artificial Analysis	—
#44	Gemini 3 Flash Preview (Reasoning)Google	89.8%	View: Artificial Analysis	—
#44	GPT-5.6 Sol (low)OpenAI	89.8%	View: Artificial Analysis	—
#44	Muse Spark 1.1 (xhigh)Meta	89.8%	View: Artificial Analysis	—
#47	Hy3Unknown	89.7%	View: Artificial Analysis	—
#48	Claude Opus 4.6 (Adaptive Reasoning, Max Effort)Anthropic	89.6%	View: Artificial Analysis	—
#48	GPT-5.6 Terra (high)OpenAI	89.6%	View: Artificial Analysis	—
#48	Kimi K2.7 CodeMoonshot AI	89.6%	View: Artificial Analysis	—
#51	GLM-5.2 (max)Z.ai	89.5%	View: Artificial Analysis	—
#51	GPT-5.6 Luna (xhigh)OpenAI	89.5%	View: Artificial Analysis	—
#51	Grok Build 0.1 0616SpaceXAI	89.5%	View: Artificial Analysis	—
#51	Inkling SmallInclusionAI	89.5%	View: Artificial Analysis	—
#55	DeepSeek V4 Flash (Reasoning, Max Effort)DeepSeek	89.4%	View: Artificial Analysis	—
#56	Qwen3.5 397B A17B (Reasoning)Qwen	89.3%	View: Artificial Analysis	—
#57	GPT-5.6 Luna (high)OpenAI	89.2%	View: Artificial Analysis	—
#57	Nex-N2-ProUnknown	89.2%	View: Artificial Analysis	—
#59	Grok 4.3 (medium)SpaceXAI	89%	View: Artificial Analysis	—
#60	Claude Opus 5 (Adaptive Reasoning, Low Effort)Anthropic	88.9%	View: Artificial Analysis	—
#61	Qwen3.6 Max PreviewQwen	88.8%	View: Artificial Analysis	—
#62	Gemini 3 Pro Preview (low)Google	88.7%	View: Artificial Analysis	—
#63	Claude Opus 4.7 (Non-reasoning, High Effort)Anthropic	88.5%	View: Artificial Analysis	—
#63	Grok 4.20 0309 (Reasoning)SpaceXAI	88.5%	View: Artificial Analysis	—
#65	Muse SparkMeta	88.4%	View: Artificial Analysis	—
#66	Qwen3.6 PlusQwen	88.2%	View: Artificial Analysis	—
#67	Kimi K2.5 (Reasoning)Moonshot AI	87.9%	View: Artificial Analysis	—
#68	Grok 4SpaceXAI	87.7%	View: Artificial Analysis	—
#69	Agnes 2.5 Pro AlphaUnknown	87.6%	View: Artificial Analysis	—
#70	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)Anthropic	87.5%	View: Artificial Analysis	—
#70	GPT-5.4 mini (xhigh)OpenAI	87.5%	View: Artificial Analysis	—
#72	MiniMax-M2.7MiniMax	87.4%	View: Artificial Analysis	—
#73	GPT-5.1 (high)OpenAI	87.3%	View: Artificial Analysis	—
#74	GPT-5.6 Terra (medium)OpenAI	87.2%	View: Artificial Analysis	—
#74	Inkling (xhigh)InclusionAI	87.2%	View: Artificial Analysis	—
#76	DeepSeek V3.2 SpecialeDeepSeek	87.1%	View: Artificial Analysis	—
#76	GPT-5.4 (low)OpenAI	87.1%	View: Artificial Analysis	—
#78	MiMo-V2-ProXiaomi	87%	View: Artificial Analysis	—
#79	Motif 3 (Beta)Unknown	86.9%	View: Artificial Analysis	—
#80	GLM-5.1 (Reasoning)Z.ai	86.8%	View: Artificial Analysis	—
#81	DeepSeek V4 Flash (Reasoning, High Effort)DeepSeek	86.7%	View: Artificial Analysis	—
#81	Hy3-preview (Reasoning)Unknown	86.7%	View: Artificial Analysis	—
#81	Nemotron 3 Ultra 550B A55B (Reasoning)NVIDIA	86.7%	View: Artificial Analysis	—
#84	Claude Opus 4.5 (Reasoning)Anthropic	86.6%	View: Artificial Analysis	—
#84	MiMo-V2.5-ProXiaomi	86.6%	View: Artificial Analysis	—
#86	GPT-5.2 (medium)OpenAI	86.4%	View: Artificial Analysis	—
#87	Qwen3 Max ThinkingQwen	86.1%	View: Artificial Analysis	—
#87	Qwen3.5 397B A17B (Non-reasoning)Qwen	86.1%	View: Artificial Analysis	—
#89	GPT-5.1 Codex (high)OpenAI	86%	View: Artificial Analysis	—
#90	GLM-4.7 (Reasoning)Z.ai	85.9%	View: Artificial Analysis	—
#90	GPT-5.6 Luna (medium)OpenAI	85.9%	View: Artificial Analysis	—
#92	Qwen3.5 27B (Reasoning)Qwen	85.8%	View: Artificial Analysis	—
#93	Gemma 4 31B (Reasoning)Google	85.7%	View: Artificial Analysis	—
#93	Qwen3.5 122B A10B (Reasoning)Qwen	85.7%	View: Artificial Analysis	—
#93	Ring-2.6-1TUnknown	85.7%	View: Artificial Analysis	—
#96	KAT Coder Pro V2Unknown	85.5%	View: Artificial Analysis	—
#96	Ling 3.0 FlashUnknown	85.5%	View: Artificial Analysis	—
#96	MiMo-V2-Omni-0327Xiaomi	85.5%	View: Artificial Analysis	—
#99	GPT-5 (high)OpenAI	85.4%	View: Artificial Analysis	—
#100	Grok 4.1 Fast (Reasoning)SpaceXAI	85.3%	View: Artificial Analysis	—
#101	MiMo-V2.5Xiaomi	84.9%	View: Artificial Analysis	—
#101	Nanbeige4.1-3BUnknown	84.9%	View: Artificial Analysis	—
#103	MiniMax-M2.5MiniMax	84.8%	View: Artificial Analysis	—
#104	GLM-5-TurboZ.ai	84.7%	View: Artificial Analysis	—
#104	Grok 4 Fast (Reasoning)SpaceXAI	84.7%	View: Artificial Analysis	—
#106	GPT-5.5 Instant (May 2026)OpenAI	84.6%	View: Artificial Analysis	—
#106	MiMo-V2-Flash (Reasoning)Xiaomi	84.6%	View: Artificial Analysis	—
#108	JT-4.1 Flash 236B A21BUnknown	84.5%	View: Artificial Analysis	—
#108	o3-proOpenAI	84.5%	View: Artificial Analysis	—
#108	Qwen3.5 35B A3B (Reasoning)Qwen	84.5%	View: Artificial Analysis	—
#111	Gemini 2.5 ProGoogle	84.4%	View: Artificial Analysis	—
#112	GPT-5.6 Terra (low)OpenAI	84.3%	View: Artificial Analysis	—
#112	Grok 4.3 (low)SpaceXAI	84.3%	View: Artificial Analysis	—
#114	GPT-5 (medium)OpenAI	84.2%	View: Artificial Analysis	—
#114	Kimi K3 (low)Moonshot AI	84.2%	View: Artificial Analysis	—
#114	Qwen3.5 27B (Non-reasoning)Qwen	84.2%	View: Artificial Analysis	—
#114	Qwen3.6 27B (Reasoning)Qwen	84.2%	View: Artificial Analysis	—
#118	Qwen3.6 35B A3B (Reasoning)Qwen	84.1%	View: Artificial Analysis	—
#119	Claude Opus 4.6 (Non-reasoning, High Effort)Anthropic	84%	View: Artificial Analysis	—
#119	DeepSeek V3.2 (Reasoning)DeepSeek	84%	View: Artificial Analysis	—
#121	GLM-5.1 (Non-reasoning)Z.ai	83.9%	View: Artificial Analysis	—
#122	Gemini 3.5 Flash-LiteGoogle	83.8%	View: Artificial Analysis	—
#122	Kimi K2 ThinkingMoonshot AI	83.8%	View: Artificial Analysis	—
#124	GPT-5 Codex (high)OpenAI	83.7%	View: Artificial Analysis	—
#125	Gemini 2.5 Pro Preview (Mar' 25)Google	83.6%	View: Artificial Analysis	—
#126	GPT-5.6 Luna (low)OpenAI	83.5%	View: Artificial Analysis	—
#126	MiMo-V2-Flash (Feb 2026)Xiaomi	83.5%	View: Artificial Analysis	—
#126	Muse Glimmer (high)Meta	83.5%	View: Artificial Analysis	—
#129	Claude 4.5 Sonnet (Reasoning)Anthropic	83.4%	View: Artificial Analysis	—
#130	Step 3.5 FlashUnknown	83.1%	View: Artificial Analysis	—
#131	MiniMax-M2.1MiniMax	83%	View: Artificial Analysis	—
#132	JT-35B-FlashUnknown	82.9%	View: Artificial Analysis	—
#132	Qwen3.6 27B (Non-reasoning)Qwen	82.9%	View: Artificial Analysis	—
#134	Gemini 3.5 Flash (minimal)Google	82.8%	View: Artificial Analysis	—
#134	GPT-5 mini (high)OpenAI	82.8%	View: Artificial Analysis	—
#134	MiMo-V2-OmniXiaomi	82.8%	View: Artificial Analysis	—
#137	o3OpenAI	82.7%	View: Artificial Analysis	—
#137	Qwen3.5 122B A10B (Non-reasoning)Qwen	82.7%	View: Artificial Analysis	—
#139	Qwen3.5 Omni PlusQwen	82.6%	View: Artificial Analysis	—
#139	Step 3.5 Flash 2603Unknown	82.6%	View: Artificial Analysis	—
#141	GPT-5.4 mini (medium)OpenAI	82.3%	View: Artificial Analysis	—
#141	GPT-5.5 Instant (June 2026)OpenAI	82.3%	View: Artificial Analysis	—
#143	Gemini 2.5 Pro Preview (May' 25)Google	82.2%	View: Artificial Analysis	—
#143	Gemini 3.1 Flash-LiteGoogle	82.2%	View: Artificial Analysis	—
#145	GLM-5 (Reasoning)Z.ai	82%	View: Artificial Analysis	—
#146	Qwen3.5 35B A3B (Non-reasoning)Qwen	81.9%	View: Artificial Analysis	—
#147	GPT-5.4 nano (xhigh)OpenAI	81.7%	View: Artificial Analysis	—
#147	Qwen3.6 35B A3B (Non-reasoning)Qwen	81.7%	View: Artificial Analysis	—
#149	DeepSeek R1 0528 (May '25)DeepSeek	81.3%	View: Artificial Analysis	—
#149	GPT-5.1 Codex mini (high)OpenAI	81.3%	View: Artificial Analysis	—
#151	Gemini 3 Flash Preview (Non-reasoning)Google	81.2%	View: Artificial Analysis	—
#152	ERNIE 4.5 300B A47BBaidu	81.1%	View: Artificial Analysis	—
#152	Nova 2.0 Lite (high)Amazon	81.1%	View: Artificial Analysis	—
#154	Claude Opus 4.5 (Non-reasoning)Anthropic	81%	View: Artificial Analysis	—
#155	Claude 4.1 Opus (Reasoning)Anthropic	80.9%	View: Artificial Analysis	—
#155	GLM 5V Turbo (Reasoning)Z.ai	80.9%	View: Artificial Analysis	—
#155	Step 3.7 FlashUnknown	80.9%	View: Artificial Analysis	—
#158	GPT-5 (low)OpenAI	80.8%	View: Artificial Analysis	—
#159	Qwen3.5 9B (Reasoning)Qwen	80.6%	View: Artificial Analysis	—
#160	GPT-5 mini (medium)OpenAI	80.3%	View: Artificial Analysis	—
#161	Claude Sonnet 5 (Non-reasoning, High Effort)Anthropic	80%	View: Artificial Analysis	—
#161	Nemotron 3 Super 120B A12B (Reasoning)NVIDIA	80%	View: Artificial Analysis	—
#163	Claude Sonnet 4.6 (Non-reasoning, High Effort)Anthropic	79.9%	View: Artificial Analysis	—
#164	Claude Sonnet 4.6 (Non-reasoning, Low Effort)Anthropic	79.7%	View: Artificial Analysis	—
#164	DeepSeek V3.2 Exp (Reasoning)DeepSeek	79.7%	View: Artificial Analysis	—
#166	Claude 4 Opus (Reasoning)Anthropic	79.6%	View: Artificial Analysis	—
#167	EXAONE 4.5 33BUpstage	79.4%	View: Artificial Analysis	—
#168	Gemini 2.5 Flash Preview (Sep '25) (Reasoning)Google	79.3%	View: Artificial Analysis	—
#169	DeepSeek V3.1 Terminus (Reasoning)DeepSeek	79.2%	View: Artificial Analysis	—
#169	Gemma 4 26B A4B (Reasoning)Google	79.2%	View: Artificial Analysis	—
#171	Grok 3 mini Reasoning (high)SpaceXAI	79.1%	View: Artificial Analysis	—
#172	Gemini 2.5 Flash (Reasoning)Google	79%	View: Artificial Analysis	—
#172	GPT-5.6 Sol (Non-reasoning)OpenAI	79%	View: Artificial Analysis	—
#172	Qwen3 235B A22B 2507 (Reasoning)Qwen	79%	View: Artificial Analysis	—
#175	Kimi K2.5 (Non-reasoning)Moonshot AI	78.9%	View: Artificial Analysis	—
#176	Kimi K2.6 (Non-reasoning)Moonshot AI	78.8%	View: Artificial Analysis	—
#177	Qwen3.5 9B (Non-reasoning)Qwen	78.6%	View: Artificial Analysis	—
#178	Grok 4.20 0309 (Non-reasoning)SpaceXAI	78.5%	View: Artificial Analysis	—
#178	Nova 2.0 Pro Preview (medium)Amazon	78.5%	View: Artificial Analysis	—
#180	o4-mini (high)OpenAI	78.4%	View: Artificial Analysis	—
#181	K-EXAONE (Reasoning)Upstage	78.3%	View: Artificial Analysis	—
#182	GLM-4.5 (Reasoning)Z.ai	78.2%	View: Artificial Analysis	—
#182	gpt-oss-120b (high)OpenAI	78.2%	View: Artificial Analysis	—
#184	GLM-4.6 (Reasoning)Z.ai	78%	View: Artificial Analysis	—
#184	LongCat 2.0Tencent	78%	View: Artificial Analysis	—
#186	DeepSeek V3.1 (Reasoning)DeepSeek	77.9%	View: Artificial Analysis	—
#187	Claude 4 Sonnet (Reasoning)Anthropic	77.7%	View: Artificial Analysis	—
#187	ERNIE 5.0 Thinking PreviewBaidu	77.7%	View: Artificial Analysis	—
#187	MiniMax-M2MiniMax	77.7%	View: Artificial Analysis	—
#190	Grok 4.20 0309 v2 (Non-reasoning)SpaceXAI	77.6%	View: Artificial Analysis	—
#190	Qwen3 Max Thinking (Preview)Qwen	77.6%	View: Artificial Analysis	—
#192	Ring-1TUnknown	77.4%	View: Artificial Analysis	—
#193	o3-mini (high)OpenAI	77.3%	View: Artificial Analysis	—
#194	Claude 3.7 Sonnet (Reasoning)Anthropic	77.2%	View: Artificial Analysis	—
#194	Qwen3 VL 235B A22B (Reasoning)Qwen	77.2%	View: Artificial Analysis	—
#196	Qwen3.5 4B (Reasoning)Qwen	77.1%	View: Artificial Analysis	—
#197	Mercury 2Inception	77%	View: Artificial Analysis	—
#198	Mistral Small 4 (Reasoning)Mistral	76.9%	View: Artificial Analysis	—
#199	Cogito v2.1 (Reasoning)Unknown	76.8%	View: Artificial Analysis	—
#199	GPT-5.5 (Non-reasoning)OpenAI	76.8%	View: Artificial Analysis	—
#199	Nova 2.0 Lite (medium)Amazon	76.8%	View: Artificial Analysis	—
#202	Kimi K2 0905Moonshot AI	76.7%	View: Artificial Analysis	—
#203	Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)Google	76.6%	View: Artificial Analysis	—
#203	Kimi K2Moonshot AI	76.6%	View: Artificial Analysis	—
#205	Doubao Seed CodeByteDance	76.4%	View: Artificial Analysis	—
#205	KAT-Coder-Pro V1Unknown	76.4%	View: Artificial Analysis	—
#205	Qwen3 MaxQwen	76.4%	View: Artificial Analysis	—
#205	Qwen3 Max (Preview)Qwen	76.4%	View: Artificial Analysis	—
#209	Gemma 4 31B (Non-reasoning)Google	76.3%	View: Artificial Analysis	—
#210	MiMo-V2.5-Pro (Non-reasoning)Xiaomi	76.2%	View: Artificial Analysis	—
#211	Command A+Cohere	76.1%	View: Artificial Analysis	—
#211	GPT-5.4 nano (medium)OpenAI	76.1%	View: Artificial Analysis	—
#211	INTELLECT-3Unknown	76.1%	View: Artificial Analysis	—
#214	Nova 2.0 Omni (medium)Amazon	76%	View: Artificial Analysis	—
#215	Qwen3 Next 80B A3B (Reasoning)Qwen	75.9%	View: Artificial Analysis	—
#216	Nemotron Cascade 2 30B A3BNVIDIA	75.8%	View: Artificial Analysis	—
#217	North Mini CodeUnknown	75.7%	View: Artificial Analysis	—
#217	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)NVIDIA	75.7%	View: Artificial Analysis	—
#219	G9v3-39A5BUnknown	75.6%	View: Artificial Analysis	—
#220	Gemma 4 12B (Reasoning)Google	75.3%	View: Artificial Analysis	—
#220	Qwen3 235B A22B 2507 InstructQwen	75.3%	View: Artificial Analysis	—
#222	Ling-2.6-1TUnknown	75.2%	View: Artificial Analysis	—
#222	Trinity Large ThinkingUnknown	75.2%	View: Artificial Analysis	—
#224	DeepSeek V3.1 Terminus (Non-reasoning)DeepSeek	75.1%	View: Artificial Analysis	—
#224	DeepSeek V3.2 (Non-reasoning)DeepSeek	75.1%	View: Artificial Analysis	—
#224	Nova 2.0 Pro Preview (low)Amazon	75.1%	View: Artificial Analysis	—
#227	GPT-5.4 (Non-reasoning)OpenAI	74.8%	View: Artificial Analysis	—
#227	Llama Nemotron Super 49B v1.5 (Reasoning)Meta	74.8%	View: Artificial Analysis	—
#227	Mistral Medium 3.5Mistral	74.8%	View: Artificial Analysis	—
#227	o3-miniOpenAI	74.8%	View: Artificial Analysis	—
#231	o1OpenAI	74.7%	View: Artificial Analysis	—
#232	GPT-5.6 Terra (Non-reasoning)OpenAI	74.6%	View: Artificial Analysis	—
#233	Nemotron 3.5 LightningNVIDIA	74.3%	View: Artificial Analysis	—
#234	Qwen3.5 Omni FlashQwen	74.2%	View: Artificial Analysis	—
#235	EXAONE 4.0 32B (Reasoning)Upstage	73.9%	View: Artificial Analysis	—
#235	Magistral Medium 1.2Mistral	73.9%	View: Artificial Analysis	—
#237	DeepSeek V3.2 Exp (Non-reasoning)DeepSeek	73.8%	View: Artificial Analysis	—
#237	Qwen3 Next 80B A3B InstructQwen	73.8%	View: Artificial Analysis	—
#237	Sarvam 105B (high)Unknown	73.8%	View: Artificial Analysis	—
#240	Qwen3 Coder NextQwen	73.7%	View: Artificial Analysis	—
#241	DeepSeek V3.1 (Non-reasoning)DeepSeek	73.5%	View: Artificial Analysis	—
#242	Ling 3.0 TinyUnknown	73.4%	View: Artificial Analysis	—
#243	Apriel-v1.6-15B-ThinkerUnknown	73.3%	View: Artificial Analysis	—
#243	GLM-4.5-AirZ.ai	73.3%	View: Artificial Analysis	—
#243	HyperNova 60B 2605Amazon	73.3%	View: Artificial Analysis	—
#243	Qwen3 VL 32B (Reasoning)Qwen	73.3%	View: Artificial Analysis	—
#247	Hy3-preview (Non-reasoning)Unknown	73.2%	View: Artificial Analysis	—
#248	Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)Meta	72.8%	View: Artificial Analysis	—
#249	Claude 4.5 Sonnet (Non-reasoning)Anthropic	72.7%	View: Artificial Analysis	—
#249	Grok Code Fast 1SpaceXAI	72.7%	View: Artificial Analysis	—
#249	Hermes 4 - Llama-3.1 405B (Reasoning)Meta	72.7%	View: Artificial Analysis	—
#252	Qwen3 Omni 30B A3B (Reasoning)Qwen	72.6%	View: Artificial Analysis	—
#252	Seed-OSS-36B-InstructByteDance	72.6%	View: Artificial Analysis	—
#254	Ring-flash-2.0Unknown	72.5%	View: Artificial Analysis	—
#255	Solar Pro 3Upstage	72.4%	View: Artificial Analysis	—
#256	Mi:dm K 2.5 Pro PreviewUnknown	72.2%	View: Artificial Analysis	—
#257	Qwen3 VL 30B A3B (Reasoning)Qwen	72%	View: Artificial Analysis	—
#258	GLM-4.6V (Reasoning)Z.ai	71.9%	View: Artificial Analysis	—
#258	Ling-1TUnknown	71.9%	View: Artificial Analysis	—
#260	DeepSeek V4 Pro (Non-reasoning)DeepSeek	71.7%	View: Artificial Analysis	—
#261	DeepSeek V4 Flash (Non-reasoning)DeepSeek	71.6%	View: Artificial Analysis	—
#262	Gemma 4 26B A4B (Non-reasoning)Google	71.4%	View: Artificial Analysis	—
#263	Apriel-v1.5-15B-ThinkerUnknown	71.3%	View: Artificial Analysis	—
#263	K2 Think V2Unknown	71.3%	View: Artificial Analysis	—
#265	GPT-5.2 (Non-reasoning)OpenAI	71.2%	View: Artificial Analysis	—
#265	Qwen3 VL 235B A22B InstructQwen	71.2%	View: Artificial Analysis	—
#265	Qwen3.5 4B (Non-reasoning)Qwen	71.2%	View: Artificial Analysis	—
#268	Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)Google	70.9%	View: Artificial Analysis	—
#269	DeepSeek R1 (Jan '25)DeepSeek	70.8%	View: Artificial Analysis	—
#270	Qwen3 30B A3B 2507 (Reasoning)Qwen	70.7%	View: Artificial Analysis	—
#271	Claude 4 Opus (Non-reasoning)Anthropic	70.1%	View: Artificial Analysis	—
#271	Gemini 2.0 Flash Thinking Experimental (Jan '25)Google	70.1%	View: Artificial Analysis	—
#271	Mi:dm K 2.5 ProUnknown	70.1%	View: Artificial Analysis	—
#274	Qwen3 235B A22B (Reasoning)Qwen	70%	View: Artificial Analysis	—
#275	Hermes 4 - Llama-3.1 70B (Reasoning)Meta	69.9%	View: Artificial Analysis	—
#275	Nova 2.0 Omni (low)Amazon	69.9%	View: Artificial Analysis	—
#277	Gemini 2.5 Flash Preview (Reasoning)Google	69.8%	View: Artificial Analysis	—
#277	Nova 2.0 Lite (low)Amazon	69.8%	View: Artificial Analysis	—
#279	MiniMax M1 80kMiniMax	69.7%	View: Artificial Analysis	—
#280	K-EXAONE (Non-reasoning)Upstage	69.5%	View: Artificial Analysis	—
#280	Motif-2-12.7B-ReasoningUnknown	69.5%	View: Artificial Analysis	—
#280	Qwen3 VL 30B A3B InstructQwen	69.5%	View: Artificial Analysis	—
#283	Grok 3SpaceXAI	69.3%	View: Artificial Analysis	—
#284	Step3 VL 10BUnknown	69%	View: Artificial Analysis	—
#285	gpt-oss-20b (high)OpenAI	68.8%	View: Artificial Analysis	—
#286	GPT-5 mini (minimal)OpenAI	68.7%	View: Artificial Analysis	—
#286	Solar Pro 2 (Reasoning)Upstage	68.7%	View: Artificial Analysis	—
#288	GLM-5.2 (Non-reasoning)Z.ai	68.6%	View: Artificial Analysis	—
#288	GPT-5 (ChatGPT)OpenAI	68.6%	View: Artificial Analysis	—
#290	GLM-4.5V (Reasoning)Z.ai	68.4%	View: Artificial Analysis	—
#291	Claude 4 Sonnet (Non-reasoning)Anthropic	68.3%	View: Artificial Analysis	—
#291	Gemini 2.5 Flash (Non-reasoning)Google	68.3%	View: Artificial Analysis	—
#293	MiniMax M1 40kMiniMax	68.2%	View: Artificial Analysis	—
#294	K2-V2 (high)Unknown	68.1%	View: Artificial Analysis	—
#295	Mistral Large 3Mistral	68%	View: Artificial Analysis	—
#296	Magistral Medium 1Mistral	67.9%	View: Artificial Analysis	—
#297	GPT-5 nano (high)OpenAI	67.6%	View: Artificial Analysis	—
#297	JT-MINIUnknown	67.6%	View: Artificial Analysis	—
#299	GPT-5 (minimal)OpenAI	67.3%	View: Artificial Analysis	—
#300	Claude 4.5 Haiku (Reasoning)Anthropic	67.2%	View: Artificial Analysis	—
#300	gpt-oss-120b (low)OpenAI	67.2%	View: Artificial Analysis	—
#302	Llama 4 MaverickMeta	67.1%	View: Artificial Analysis	—
#302	Qwen3 VL 32B InstructQwen	67.1%	View: Artificial Analysis	—
#304	GPT-5 nano (medium)OpenAI	67%	View: Artificial Analysis	—
#305	DiffusionGemma 26B A4BGoogle	66.9%	View: Artificial Analysis	—
#306	Qwen3 32B (Reasoning)Qwen	66.8%	View: Artificial Analysis	—
#307	Qwen3 4B 2507 (Reasoning)Qwen	66.7%	View: Artificial Analysis	—
#308	GLM-5 (Non-reasoning)Z.ai	66.6%	View: Artificial Analysis	—
#308	GPT-4.1OpenAI	66.6%	View: Artificial Analysis	—
#310	GLM-4.7 (Non-reasoning)Z.ai	66.4%	View: Artificial Analysis	—
#310	GPT-4.1 miniOpenAI	66.4%	View: Artificial Analysis	—
#312	Magistral Small 1.2Mistral	66.3%	View: Artificial Analysis	—
#313	Falcon-H1R-7BUnknown	66.1%	View: Artificial Analysis	—
#313	Gemma 4 12B (Non-reasoning)Google	66.1%	View: Artificial Analysis	—
#315	Qwen3 30B A3B 2507 InstructQwen	65.9%	View: Artificial Analysis	—
#316	Grok 4.3 (Non-reasoning)SpaceXAI	65.8%	View: Artificial Analysis	—
#317	Ling-flash-2.0Unknown	65.7%	View: Artificial Analysis	—
#317	Solar Open 100B (Reasoning)Upstage	65.7%	View: Artificial Analysis	—
#319	Claude 3.7 Sonnet (Non-reasoning)Anthropic	65.6%	View: Artificial Analysis	—
#319	MiMo-V2-Flash (Non-reasoning)Xiaomi	65.6%	View: Artificial Analysis	—
#321	DeepSeek V3 0324DeepSeek	65.5%	View: Artificial Analysis	—
#321	GPT-4o (March 2025, chatgpt-4o-latest)OpenAI	65.5%	View: Artificial Analysis	—
#323	Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)Google	65.1%	View: Artificial Analysis	—
#324	Claude 4.5 Haiku (Non-reasoning)Anthropic	64.6%	View: Artificial Analysis	—
#325	GPT-5.6 Luna (Non-reasoning)OpenAI	64.5%	View: Artificial Analysis	—
#326	GPT-5.1 (Non-reasoning)OpenAI	64.3%	View: Artificial Analysis	—
#326	Llama 3.3 Nemotron Super 49B v1 (Reasoning)Meta	64.3%	View: Artificial Analysis	—
#328	Magistral Small 1Mistral	64.1%	View: Artificial Analysis	—
#329	Grok 4.1 Fast (Non-reasoning)SpaceXAI	63.7%	View: Artificial Analysis	—
#330	Gemini 2.0 Flash (experimental)Google	63.6%	View: Artificial Analysis	—
#330	LongCat Flash LiteTencent	63.6%	View: Artificial Analysis	—
#330	Nova 2.0 Pro Preview (Non-reasoning)Amazon	63.6%	View: Artificial Analysis	—
#333	Sarvam 30B (high)Unknown	63.3%	View: Artificial Analysis	—
#334	GLM-4.6 (Non-reasoning)Z.ai	63.2%	View: Artificial Analysis	—
#335	Celeris-1Unknown	63.1%	View: Artificial Analysis	—
#336	EXAONE 4.0 32B (Non-reasoning)Upstage	62.8%	View: Artificial Analysis	—
#337	Gemini 2.5 Flash-Lite (Reasoning)Google	62.5%	View: Artificial Analysis	—
#338	Gemini 2.0 Flash (Feb '25)Google	62.3%	View: Artificial Analysis	—
#338	Sonar ReasoningPerplexity	62.3%	View: Artificial Analysis	—
#340	Gemini 2.0 Pro Experimental (Feb '25)Google	62.2%	View: Artificial Analysis	—
#341	Qwen3 Omni 30B A3B InstructQwen	62%	View: Artificial Analysis	—
#342	Qwen3 Coder 480B A35B InstructQwen	61.8%	View: Artificial Analysis	—
#343	Qwen3 30B A3B (Reasoning)Qwen	61.6%	View: Artificial Analysis	—
#344	DeepSeek R1 Distill Qwen 32BQwen	61.5%	View: Artificial Analysis	—
#344	HyperCLOVA X SEED Think (32B)ByteDance	61.5%	View: Artificial Analysis	—
#346	Qwen3 235B A22B (Non-reasoning)Qwen	61.3%	View: Artificial Analysis	—
#347	DeepSeek R1 0528 Qwen3 8BQwen	61.2%	View: Artificial Analysis	—
#348	gpt-oss-20b (low)OpenAI	61.1%	View: Artificial Analysis	—
#349	Olmo 3 32B ThinkAllen Institute for AI	61%	View: Artificial Analysis	—
#350	GPT-5.4 mini (Non-Reasoning)OpenAI	60.6%	View: Artificial Analysis	—
#350	Grok 4 Fast (Non-reasoning)SpaceXAI	60.6%	View: Artificial Analysis	—
#352	Qwen3 14B (Reasoning)Qwen	60.4%	View: Artificial Analysis	—
#353	Nova 2.0 Lite (Non-reasoning)Amazon	60.3%	View: Artificial Analysis	—
#353	o1-miniOpenAI	60.3%	View: Artificial Analysis	—
#355	Tri-21B-ThinkUnknown	60.1%	View: Artificial Analysis	—
#356	Claude 3.5 Sonnet (Oct '24)Anthropic	59.9%	View: Artificial Analysis	—
#357	K2-V2 (medium)Unknown	59.8%	View: Artificial Analysis	—
#358	Devstral 2Mistral	59.4%	View: Artificial Analysis	—
#358	Gemini 2.5 Flash Preview (Non-reasoning)Google	59.4%	View: Artificial Analysis	—
#360	Ling 2.6 FlashUnknown	59.3%	View: Artificial Analysis	—
#360	QwQ 32BQwen	59.3%	View: Artificial Analysis	—
#362	Olmo 3.1 32B ThinkAllen Institute for AI	59.1%	View: Artificial Analysis	—
#363	Gemini 1.5 Pro (Sep '24)Google	58.9%	View: Artificial Analysis	—
#363	Qwen3 8B (Reasoning)Qwen	58.9%	View: Artificial Analysis	—
#365	Mistral Medium 3.1Mistral	58.8%	View: Artificial Analysis	—
#366	Llama 4 ScoutMeta	58.7%	View: Artificial Analysis	—
#366	Qwen2.5 MaxQwen	58.7%	View: Artificial Analysis	—
#368	GLM-4.7-Flash (Reasoning)Z.ai	58.1%	View: Artificial Analysis	—
#369	Qwen3 VL 8B (Reasoning)Qwen	57.9%	View: Artificial Analysis	—
#370	Mistral Medium 3Mistral	57.8%	View: Artificial Analysis	—
#370	Solar Pro 2 (Preview) (Reasoning)Upstage	57.8%	View: Artificial Analysis	—
#370	Sonar ProPerplexity	57.8%	View: Artificial Analysis	—
#373	Gemma 4 E4B (Reasoning)Google	57.6%	View: Artificial Analysis	—
#374	Phi-4Microsoft	57.5%	View: Artificial Analysis	—
#375	GLM-4.5V (Non-reasoning)Z.ai	57.3%	View: Artificial Analysis	—
#376	Ministral 3 14BMistral	57.2%	View: Artificial Analysis	—
#376	NVIDIA Nemotron Nano 12B v2 VL (Reasoning)NVIDIA	57.2%	View: Artificial Analysis	—
#378	Mistral Small 4 (Non-reasoning)Mistral	57.1%	View: Artificial Analysis	—
#379	NVIDIA Nemotron Nano 9B V2 (Reasoning)NVIDIA	57%	View: Artificial Analysis	—
#380	Nova PremierAmazon	56.9%	View: Artificial Analysis	—
#381	GLM-4.6V (Non-reasoning)Z.ai	56.6%	View: Artificial Analysis	—
#382	Ling-mini-2.0Unknown	56.2%	View: Artificial Analysis	—
#383	Solar Pro 2 (Non-reasoning)Upstage	56.1%	View: Artificial Analysis	—
#384	Claude 3.5 Sonnet (June '24)Anthropic	56%	View: Artificial Analysis	—
#385	GPT-5.4 nano (Non-Reasoning)OpenAI	55.8%	View: Artificial Analysis	—
#386	DeepSeek V3 (Dec '24)DeepSeek	55.7%	View: Artificial Analysis	—
#386	NVIDIA Nemotron Nano 9B V2 (Non-reasoning)NVIDIA	55.7%	View: Artificial Analysis	—
#386	QwQ 32B-PreviewQwen	55.7%	View: Artificial Analysis	—
#389	Nova 2.0 Omni (Non-reasoning)Amazon	55.5%	View: Artificial Analysis	—
#390	Gemma 4 E4B (Non-reasoning)Google	54.9%	View: Artificial Analysis	—
#391	Solar Pro 2 (Preview) (Non-reasoning)Upstage	54.4%	View: Artificial Analysis	—
#392	GPT-4o (Nov '24)OpenAI	54.3%	View: Artificial Analysis	—
#393	Gemini 2.0 Flash-Lite (Preview)Google	54.2%	View: Artificial Analysis	—
#394	K2-V2 (low)Unknown	54.1%	View: Artificial Analysis	—
#395	Olmo 3.1 32B InstructAllen Institute for AI	53.9%	View: Artificial Analysis	—
#396	Tri-21B-think PreviewUnknown	53.8%	View: Artificial Analysis	—
#397	Hermes 4 - Llama-3.1 405B (Non-reasoning)Meta	53.6%	View: Artificial Analysis	—
#398	Gemini 2.0 Flash-Lite (Feb '25)Google	53.5%	View: Artificial Analysis	—
#398	Qwen3 32B (Non-reasoning)Qwen	53.5%	View: Artificial Analysis	—
#400	Devstral Small 2Mistral	53.2%	View: Artificial Analysis	—
#401	Reka Flash 3Reka AI	52.9%	View: Artificial Analysis	—
#402	Command ACohere	52.7%	View: Artificial Analysis	—
#403	GPT-4o (May '24)OpenAI	52.6%	View: Artificial Analysis	—
#404	Qwen3 4B (Reasoning)Qwen	52.2%	View: Artificial Analysis	—
#405	GPT-4o (Aug '24)OpenAI	52.1%	View: Artificial Analysis	—
#406	Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)Meta	51.7%	View: Artificial Analysis	—
#406	Qwen3 4B 2507 InstructQwen	51.7%	View: Artificial Analysis	—
#408	Llama 3.1 Tulu3 405BMeta	51.6%	View: Artificial Analysis	—
#408	Olmo 3 7B ThinkAllen Institute for AI	51.6%	View: Artificial Analysis	—
#408	Qwen3 Coder 30B A3B InstructQwen	51.6%	View: Artificial Analysis	—
#411	Exaone 4.0 1.2B (Reasoning)Upstage	51.5%	View: Artificial Analysis	—
#411	Llama 3.1 Instruct 405BMeta	51.5%	View: Artificial Analysis	—
#411	Qwen3 30B A3B (Non-reasoning)Qwen	51.5%	View: Artificial Analysis	—
#414	LFM2.5-8B-A1BLiquid AI	51.3%	View: Artificial Analysis	—
#414	NVIDIA Nemotron 3 Nano 4BNVIDIA	51.3%	View: Artificial Analysis	—
#416	GPT-4.1 nanoOpenAI	51.2%	View: Artificial Analysis	—
#417	GPT-4o (ChatGPT)OpenAI	51.1%	View: Artificial Analysis	—
#418	Grok 2 (Dec '24)SpaceXAI	51%	View: Artificial Analysis	—
#419	Mistral Small 3.2Mistral	50.5%	View: Artificial Analysis	—
#419	Pixtral LargeUnknown	50.5%	View: Artificial Analysis	—
#421	Nova ProAmazon	49.9%	View: Artificial Analysis	—
#422	Llama 3.3 Instruct 70BMeta	49.8%	View: Artificial Analysis	—
#423	Qwen3 VL 4B (Reasoning)Qwen	49.4%	View: Artificial Analysis	—
#424	Devstral MediumMistral	49.2%	View: Artificial Analysis	—
#425	Hermes 4 - Llama-3.1 70B (Non-reasoning)Meta	49.1%	View: Artificial Analysis	—
#425	Qwen2.5 Instruct 72BQwen	49.1%	View: Artificial Analysis	—
#427	Claude 3 OpusAnthropic	48.9%	View: Artificial Analysis	—
#428	Mistral Large 2 (Nov '24)Mistral	48.6%	View: Artificial Analysis	—
#429	DeepSeek R1 Distill Qwen 14BQwen	48.4%	View: Artificial Analysis	—
#430	Granite 4.1 30BIBM	48.1%	View: Artificial Analysis	—
#430	Llama Nemotron Super 49B v1.5 (Non-reasoning)Meta	48.1%	View: Artificial Analysis	—
#432	Gemini 2.5 Flash-Lite (Non-reasoning)Google	47.4%	View: Artificial Analysis	—
#432	LFM2 24B A2BLiquid AI	47.4%	View: Artificial Analysis	—
#434	Mistral Large 2 (Jul '24)Mistral	47.2%	View: Artificial Analysis	—
#435	Grok BetaSpaceXAI	47.1%	View: Artificial Analysis	—
#435	Ministral 3 8BMistral	47.1%	View: Artificial Analysis	—
#435	SonarPerplexity	47.1%	View: Artificial Analysis	—
#438	Qwen3 14B (Non-reasoning)Qwen	47%	View: Artificial Analysis	—
#439	Nemotron 3 Nano Omni 30B A3B ReasoningNVIDIA	46.9%	View: Artificial Analysis	—
#440	Qwen2.5 Instruct 32BQwen	46.6%	View: Artificial Analysis	—
#441	Llama 3.1 Nemotron Instruct 70BMeta	46.5%	View: Artificial Analysis	—
#442	Gemini 1.5 Flash (Sep '24)Google	46.3%	View: Artificial Analysis	—
#443	Mistral Small 3Mistral	46.2%	View: Artificial Analysis	—
#444	Qwen3.5 2B (Reasoning)Qwen	45.6%	View: Artificial Analysis	—
#445	Mistral Small 3.1Mistral	45.4%	View: Artificial Analysis	—
#446	GLM-4.7-Flash (Non-reasoning)Z.ai	45.2%	View: Artificial Analysis	—
#446	Qwen3 8B (Non-reasoning)Qwen	45.2%	View: Artificial Analysis	—
#448	NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)NVIDIA	43.9%	View: Artificial Analysis	—
#449	G9v3-3BUnknown	43.8%	View: Artificial Analysis	—
#449	Qwen3.5 2B (Non-reasoning)Qwen	43.8%	View: Artificial Analysis	—
#451	Devstral Small (May '25)Mistral	43.4%	View: Artificial Analysis	—
#452	Gemma 4 E2B (Reasoning)Google	43.3%	View: Artificial Analysis	—
#452	Granite 4.1 8BIBM	43.3%	View: Artificial Analysis	—
#452	Nova LiteAmazon	43.3%	View: Artificial Analysis	—
#455	Llama 3.2 Instruct 90B (Vision)Meta	43.2%	View: Artificial Analysis	—
#456	Gemma 3 27B InstructGoogle	42.8%	View: Artificial Analysis	—
#456	GPT-5 nano (minimal)OpenAI	42.8%	View: Artificial Analysis	—
#458	Jamba 1.5 LargeAI21 Labs	42.7%	View: Artificial Analysis	—
#458	Qwen3 VL 8B InstructQwen	42.7%	View: Artificial Analysis	—
#460	GPT-4o miniOpenAI	42.6%	View: Artificial Analysis	—
#461	Molmo2-8BAllen Institute for AI	42.5%	View: Artificial Analysis	—
#462	Exaone 4.0 1.2B (Non-reasoning)Upstage	42.4%	View: Artificial Analysis	—
#462	Mistral SabaMistral	42.4%	View: Artificial Analysis	—
#464	Qwen2.5 Coder Instruct 32BQwen	41.7%	View: Artificial Analysis	—
#465	Granite 4.0 H SmallIBM	41.6%	View: Artificial Analysis	—
#465	Sarvam M (Reasoning)Unknown	41.6%	View: Artificial Analysis	—
#467	Devstral Small (Jul '25)Mistral	41.4%	View: Artificial Analysis	—
#468	Kimi Linear 48B A3B InstructMoonshot AI	41.2%	View: Artificial Analysis	—
#469	Qwen2.5 TurboQwen	41%	View: Artificial Analysis	—
#470	Llama 3.1 Instruct 70BMeta	40.9%	View: Artificial Analysis	—
#471	Claude 3.5 HaikuAnthropic	40.8%	View: Artificial Analysis	—
#471	Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)Meta	40.8%	View: Artificial Analysis	—
#473	Gemma 4 E2B (Non-reasoning)Google	40.5%	View: Artificial Analysis	—
#474	DeepSeek R1 Distill Llama 70BDeepSeek	40.2%	View: Artificial Analysis	—
#475	Hermes 3 - Llama-3.1 70BMeta	40.1%	View: Artificial Analysis	—
#476	Claude 3 SonnetAnthropic	40%	View: Artificial Analysis	—
#476	Olmo 3 7B InstructAllen Institute for AI	40%	View: Artificial Analysis	—
#478	NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)NVIDIA	39.9%	View: Artificial Analysis	—
#479	Qwen3 4B (Non-reasoning)Qwen	39.8%	View: Artificial Analysis	—
#480	Jamba 1.7 LargeAI21 Labs	39%	View: Artificial Analysis	—
#481	Jamba 1.6 LargeAI21 Labs	38.7%	View: Artificial Analysis	—
#482	DeepHermes 3 - Mistral 24B Preview (Non-reasoning)Mistral	38.2%	View: Artificial Analysis	—
#483	Mistral Small (Sep '24)Mistral	38.1%	View: Artificial Analysis	—
#484	Llama 3 Instruct 70BMeta	37.9%	View: Artificial Analysis	—
#485	Claude 3 HaikuAnthropic	37.4%	View: Artificial Analysis	—
#486	Gemini 1.5 Pro (May '24)Google	37.1%	View: Artificial Analysis	—
#486	Qwen2 Instruct 72BQwen	37.1%	View: Artificial Analysis	—
#486	Qwen3 VL 4B InstructQwen	37.1%	View: Artificial Analysis	—
#489	Gemini 1.5 Flash-8BGoogle	35.9%	View: Artificial Analysis	—
#490	Ministral 3 3BMistral	35.8%	View: Artificial Analysis	—
#490	Nova MicroAmazon	35.8%	View: Artificial Analysis	—
#492	Qwen3 1.7B (Reasoning)Qwen	35.6%	View: Artificial Analysis	—
#493	Mistral Large (Feb '24)Mistral	35.1%	View: Artificial Analysis	—
#494	Gemma 3 12B InstructGoogle	34.9%	View: Artificial Analysis	—
#494	GPT-4OpenAI	34.9%	View: Artificial Analysis	—
#494	Mistral MediumMistral	34.9%	View: Artificial Analysis	—
#497	Claude 2.0Anthropic	34.4%	View: Artificial Analysis	—
#497	LFM2 8B A1BLiquid AI	34.4%	View: Artificial Analysis	—
#499	LFM2.5-1.2B-ThinkingLiquid AI	33.9%	View: Artificial Analysis	—
#499	Qwen2.5 Coder Instruct 7B Qwen	33.9%	View: Artificial Analysis	—
#501	Granite 3.3 8B (Non-reasoning)IBM	33.8%	View: Artificial Analysis	—
#502	Granite 4.0 MicroIBM	33.6%	View: Artificial Analysis	—
#503	Jamba Reasoning 3BAI21 Labs	33.3%	View: Artificial Analysis	—
#504	Mixtral 8x22B InstructMistral	33.2%	View: Artificial Analysis	—
#505	DBRX InstructUnknown	33.1%	View: Artificial Analysis	—
#505	Phi-4 Mini InstructMicrosoft	33.1%	View: Artificial Analysis	—
#507	Claude InstantAnthropic	33%	View: Artificial Analysis	—
#508	OLMo 2 32BAllen Institute for AI	32.8%	View: Artificial Analysis	—
#509	LFM 40BLiquid AI	32.7%	View: Artificial Analysis	—
#509	Llama 2 Chat 70BMeta	32.7%	View: Artificial Analysis	—
#511	LFM2.5-1.2B-InstructLiquid AI	32.6%	View: Artificial Analysis	—
#512	Gemini 1.5 Flash (May '24)Google	32.4%	View: Artificial Analysis	—
#513	Command-R+ (Apr '24)Cohere	32.3%	View: Artificial Analysis	—
#514	Jamba 1.7 MiniAI21 Labs	32.2%	View: Artificial Analysis	—
#515	Llama 2 Chat 13BMeta	32.1%	View: Artificial Analysis	—
#516	Claude 2.1Anthropic	31.9%	View: Artificial Analysis	—
#516	DeepSeek Coder V2 Lite InstructDeepSeek	31.9%	View: Artificial Analysis	—
#516	Phi-3 Mini Instruct 3.8BMicrosoft	31.9%	View: Artificial Analysis	—
#519	Phi-4 Multimodal InstructMicrosoft	31.5%	View: Artificial Analysis	—
#520	Granite 4.1 3BIBM	31.4%	View: Artificial Analysis	—
#521	LFM2 2.6BLiquid AI	30.6%	View: Artificial Analysis	—
#522	MiniCPM-V 4.6 1.3BUnknown	30.5%	View: Artificial Analysis	—
#522	Tiny Aya GlobalCohere	30.5%	View: Artificial Analysis	—
#524	DeepSeek R1 Distill Llama 8BDeepSeek	30.2%	View: Artificial Analysis	—
#524	Jamba 1.5 MiniAI21 Labs	30.2%	View: Artificial Analysis	—
#524	Mistral Small (Feb '24)Mistral	30.2%	View: Artificial Analysis	—
#527	Jamba 1.6 MiniAI21 Labs	30%	View: Artificial Analysis	—
#528	GPT-3.5 TurboOpenAI	29.7%	View: Artificial Analysis	—
#529	Gemma 3n E4B InstructGoogle	29.6%	View: Artificial Analysis	—
#529	Llama 3 Instruct 8BMeta	29.6%	View: Artificial Analysis	—
#531	Mixtral 8x7B InstructMistral	29.2%	View: Artificial Analysis	—
#532	Gemma 3 4B InstructGoogle	29.1%	View: Artificial Analysis	—
#533	LFM2.5-VL-1.6BLiquid AI	28.9%	View: Artificial Analysis	—
#533	Qwen1.5 Chat 110BQwen	28.9%	View: Artificial Analysis	—
#535	OLMo 2 7BAllen Institute for AI	28.8%	View: Artificial Analysis	—
#536	Command-R (Mar '24)Cohere	28.4%	View: Artificial Analysis	—
#537	Qwen3 1.7B (Non-reasoning)Qwen	28.3%	View: Artificial Analysis	—
#538	Granite 4.0 1BIBM	28.1%	View: Artificial Analysis	—
#539	Gemma 3n E4B Instruct Preview (May '25)Google	27.8%	View: Artificial Analysis	—
#539	MiniCPM5-1B (Reasoning)Unknown	27.8%	View: Artificial Analysis	—
#541	Gemini 1.0 ProGoogle	27.7%	View: Artificial Analysis	—
#542	Apertus 70B InstructUnknown	27.2%	View: Artificial Analysis	—
#543	DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)Meta	27%	View: Artificial Analysis	—
#544	MiniCPM5-1B (Non-reasoning)Unknown	26.9%	View: Artificial Analysis	—
#545	Granite 4.0 H 1BIBM	26.3%	View: Artificial Analysis	—
#546	Granite 4.0 350MIBM	26.1%	View: Artificial Analysis	—
#547	Llama 3.1 Instruct 8BMeta	25.9%	View: Artificial Analysis	—
#548	Granite 4.0 H 350MIBM	25.7%	View: Artificial Analysis	—
#549	Apertus 8B InstructUnknown	25.6%	View: Artificial Analysis	—
#550	Llama 3.2 Instruct 3BMeta	25.5%	View: Artificial Analysis	—
#551	Molmo 7B-DAllen Institute for AI	24%	View: Artificial Analysis	—
#552	Qwen3 0.6B (Reasoning)Qwen	23.9%	View: Artificial Analysis	—
#553	Gemma 3 1B InstructGoogle	23.7%	View: Artificial Analysis	—
#554	Qwen3.5 0.8B (Non-reasoning)Qwen	23.6%	View: Artificial Analysis	—
#555	Qwen3 0.6B (Non-reasoning)Qwen	23.1%	View: Artificial Analysis	—
#556	OpenChat 3.5 (1210)Unknown	23%	View: Artificial Analysis	—
#557	Gemma 3n E2B InstructGoogle	22.9%	View: Artificial Analysis	—
#558	LFM2 1.2BLiquid AI	22.8%	View: Artificial Analysis	—
#559	Llama 2 Chat 7BMeta	22.7%	View: Artificial Analysis	—
#560	Gemma 3 270MGoogle	22.4%	View: Artificial Analysis	—
#561	Llama 3.2 Instruct 11B (Vision)Meta	22.1%	View: Artificial Analysis	—
#562	Llama 3.2 Instruct 1BMeta	19.6%	View: Artificial Analysis	—
#563	Mistral 7B InstructMistral	17.7%	View: Artificial Analysis	—
#564	Qwen3.5 0.8B (Reasoning)Qwen	11.1%	View: Artificial Analysis	—
#565	DeepSeek R1 Distill Qwen 1.5BQwen	9.8%	View: Artificial Analysis	—

Full Artificial Analysis GPQA Diamond table mirrored 2026-08-11 from artificialanalysis.ai, with Grok 4.6 (high) 94.9% inserted from the live AA table on 2026-08-14. GPQA Diamond has no single first-party live leaderboard; the original paper defines the 198-question split.

Open Artificial Analysis GPQA Diamond leaderboard

Evidence in this catalog

Where GPQA Diamond fits

Closest verified examples we currently carry. A missing score is not a zero.

Model examples

No directly comparable model score is verified here yet.

Harness examples

No directly comparable harness score is verified here yet.

Open this board on Artificial Analysis GPQA Diamond leaderboard