Skip to content

Leaderboards

Benchmark leaderboards

99 benchmarks across 577 scored entries — each leaderboard pairs scores with current best prices and links back to the official benchmark. Also available as JSON.

SWE-Bench Pro

51 scored · 22 independently measured · resolve rate

Measured leader

Independent
GPT-5.4

OpenAI · Independent · labs.scale.com

Score

59.1

+7.2 vs. #2

Current API price

$0.75/$6 /M

1,704 pts/$

Terminal-Bench

37 scored · 9 independently measured · success rate

Measured leader

Independent
GPT-5.5

OpenAI · Independent · artificialanalysis.ai

Score

84.1

+10.3 vs. #2

Current API price

$0.188/$1.13 /M

10,240 pts/$

SciCode

30 scored · 0 independently measured · percent correct

Top reported

No independent ranking
Fugu

Sakana · Unclassified source · console.sakana.ai

Score

60.1

Reported, not ranked

Current API price

No published price

Aider Polyglot

28 scored · 28 independently measured · percent correct

Measured leader

Independent
GPT-5

OpenAI · Independent · aider.chat

Score

88

+3.1 vs. #2

Current API price

$0.2/$1.20 /M

11,992 pts/$

Terminal-Bench Hard

22 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-5-Codex

OpenAI · Marketplace · openrouter.ai

Score

37.9

Reported, not ranked

Current API price

$1.07/$8.50 /M

771 pts/$

Humanity's Last Exam

22 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

64.7

Reported, not ranked

Current API price

$5/$25 /M

261 pts/$

GPQA Diamond

20 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

96

Reported, not ranked

Current API price

$10/$50 /M

194 pts/$

BrowseComp

15 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

91.5

Reported, not ranked

Current API price

$10/$50 /M

293 pts/$

SWE-Atlas Codebase QnA

14 scored · 14 independently measured · score

Measured leader

Independent
Claude Opus 4.7

Anthropic · Independent · artificialanalysis.ai

Score

81

+0.2 vs. #2

Current API price

$4/$20 /M

90.5 pts/$

DeepSWE

13 scored · 0 independently measured · resolve rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

74.1

Reported, not ranked

Current API price

$10/$50 /M

237 pts/$

MCP Atlas

11 scored · 0 independently measured · success rate

Top reported

No independent ranking
Muse Spark 1.1

Meta · Unclassified source · ai.meta.com

Score

88.1

Reported, not ranked

Current API price

$1.25/$4.25 /M

3,011 pts/$

SWE-Atlas Refactoring

11 scored · 11 independently measured · score

Measured leader

Independent
Claude Opus 4.7

Anthropic · Independent · labs.scale.com

Score

48.57

+3.78 vs. #2

Current API price

$4/$20 /M

54.3 pts/$

MMMU Pro

10 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Gemini 3.5 Flash

Google DeepMind · Unclassified source · deepmind.google

Score

83.6

Reported, not ranked

Current API price

$0.186/$1.11 /M

7,630 pts/$

SWE-Atlas Test Writing

10 scored · 10 independently measured · score

Measured leader

Independent
GPT-5.4

OpenAI · Independent · labs.scale.com

Score

44.36

+1.77 vs. #2

Current API price

$0.75/$6 /M

265 pts/$

GDPval-AA

9 scored · 2 independently measured · Elo

Measured leader

Independent
Grok 4.3

xAI · Independent · artificialanalysis.ai

Score

1,500

+1,486 vs. #2

Current API price

$1.25/$2.50 /M

533,808 pts/$

CharXiv Reasoning

9 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

91.3

Reported, not ranked

Current API price

$2/$8 /M

1,141 pts/$

Toolathlon

8 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-5.6 Sol

OpenAI · Unclassified source · openai.com

Score

58

Reported, not ranked

Current API price

$2/$10 /M

915 pts/$

FrontierMath

8 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

97.6

Reported, not ranked

Current API price

$10/$50 /M

197 pts/$

AutomationBench

7 scored · 0 independently measured · pass@1

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

41.4

Reported, not ranked

Current API price

$10/$50 /M

133 pts/$

FrontierCode

6 scored · 0 independently measured · pass rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

64.5

Reported, not ranked

Current API price

$10/$50 /M

130 pts/$

ARC-AGI-2

6 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

90.4

Reported, not ranked

Current API price

$5/$25 /M

365 pts/$

GDPval

6 scored · 0 independently measured · wins or ties

Top reported

No independent ranking
GPT-5.5

OpenAI · Unclassified source · openai.com

Score

84.9

Reported, not ranked

Current API price

$0.188/$1.13 /M

98,973 pts/$

OSWorld

5 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

72.6

Reported, not ranked

Current API price

$10/$50 /M

1,908 pts/$

τ²-Bench Telecom

5 scored · 2 independently measured · success rate

Measured leader

Independent
Grok 4.3

xAI · Independent · artificialanalysis.ai

Score

98

+61 vs. #2

Current API price

$1.25/$2.50 /M

34,875 pts/$

Agents' Last Exam

5 scored · 0 independently measured · score

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

59.3

Reported, not ranked

Current API price

$10/$50 /M

1,558 pts/$

NL2Repo

4 scored · 0 independently measured

Top reported

No independent ranking
Qwen3.8 Max Preview

Alibaba · Unclassified source · alibabacloud.com

Score

55.9

Reported, not ranked

Current API price

$2/$6 /M

1,303 pts/$

FrontierSWE

3 scored · 0 independently measured · dominance score

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

81.2

Reported, not ranked

Current API price

$2/$8 /M

12,257 pts/$

Toolathlon Verified

3 scored · 0 independently measured · pass@1

Top reported

No independent ranking
Muse Spark 1.1

Meta · Unclassified source · ai.meta.com

Score

75.6

Reported, not ranked

Current API price

$1.25/$4.25 /M

20,086 pts/$

IFBench

3 scored · 1 independently measured · score

Measured leader

Independent
Grok 4.3

xAI · Independent · artificialanalysis.ai

Score

81

Only independent score

Current API price

$1.25/$2.50 /M

28,826 pts/$

GDM-MRCR

3 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Gemini 3.7 Flash

Google DeepMind · Unclassified source · deepmind.google

Score

97

Reported, not ranked

Current API price

$0.75/$3.75 /M

667 pts/$

Program Bench

3 scored · 0 independently measured

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

77.8

Reported, not ranked

Current API price

$2/$8 /M

11,743 pts/$

SWE Marathon

3 scored · 0 independently measured · resolve rate

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

42

Reported, not ranked

Current API price

$2/$8 /M

6,340 pts/$

LiveCodeBench

3 scored · 0 independently measured · pass@1

Top reported

No independent ranking
Fugu Ultra

Sakana · Unclassified source · console.sakana.ai

Score

93.2

Reported, not ranked

Current API price

$5/$30 /M

4,330 pts/$

MLS-Bench-Lite

2 scored · 0 independently measured · score

Top reported

No independent ranking
Qwen3.8 Max Preview

Alibaba · Unclassified source · alibabacloud.com

Score

41

Reported, not ranked

Current API price

$2/$6 /M

7,289 pts/$

AA-Briefcase

2 scored · 0 independently measured · Elo

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

1,720

Reported, not ranked

Current API price

$5/$25 /M

89,964 pts/$

ARC-AGI-1

2 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

97.5

Reported, not ranked

Current API price

$5/$25 /M

5,100 pts/$

ARC-AGI-3

2 scored · 0 independently measured · RHAE

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

99.9

Reported, not ranked

Current API price

$10/$50 /M

2,625 pts/$

MLE-Bench

2 scored · 0 independently measured · average position score

Top reported

No independent ranking
Gemini 3.6 Flash

Google DeepMind · Unclassified source · deepmind.google

Score

63.9

Reported, not ranked

Current API price

$0.375/$1.88 /M

45,633 pts/$

JobBench

2 scored · 0 independently measured · success rate

Top reported

No independent ranking
Muse Spark 1.1

Meta · Unclassified source · ai.meta.com

Score

54.7

Reported, not ranked

Current API price

$1.25/$4.25 /M

14,533 pts/$

MMLU-Pro

2 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

90.33

Reported, not ranked

Current API price

$0.95/$4 /M

27,740 pts/$

OmniDocBench

2 scored · 0 independently measured · overall edit distance

Top reported

No independent ranking
GPT-5.4 nano

OpenAI · Unclassified source · openai.com

Score

0.24

Reported, not ranked

Current API price

$0.18/$1.10 /M

308 pts/$

OpenAI MRCR

2 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-5.4 mini

OpenAI · Unclassified source · openai.com

Score

47.7

Reported, not ranked

Current API price

$0.375/$4 /M

19,161 pts/$

Graphwalks

2 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-5.4 mini

OpenAI · Unclassified source · openai.com

Score

76.3

Reported, not ranked

Current API price

$0.375/$4 /M

30,650 pts/$

GeneBench

2 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-5.5 Pro

OpenAI · Unclassified source · openai.com

Score

33.2

Reported, not ranked

Current API price

$27.27/$163.64 /M

266 pts/$

Terminal Bench 2.1

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu Ultra

Sakana · Unclassified source · console.sakana.ai

Score

82.1

Reported, not ranked

Current API price

$5/$30 /M

453 pts/$

LiveCodeBench Pro

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu Ultra

Sakana · Unclassified source · console.sakana.ai

Score

90.8

Reported, not ranked

Current API price

$5/$30 /M

4,218 pts/$

τ3 Banking

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu

Sakana · Unclassified source · console.sakana.ai

Score

21.7

Reported, not ranked

Current API price

No published price

Long Context Reasoning

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu

Sakana · Unclassified source · console.sakana.ai

Score

74.7

Reported, not ranked

Current API price

No published price

MRCRv2

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu Ultra

Sakana · Unclassified source · console.sakana.ai

Score

93.6

Reported, not ranked

Current API price

$5/$30 /M

4,348 pts/$

CTI-REALM

2 scored · 0 independently measured

Top reported

No independent ranking
Fugu Ultra

Sakana · Unclassified source · console.sakana.ai

Score

69.4

Reported, not ranked

Current API price

$5/$30 /M

3,224 pts/$

WideSearch

1 scored · 0 independently measured · F1

Top reported

No independent ranking
Qwen3.8 Max Preview

Alibaba · Unclassified source · alibabacloud.com

Score

81.9

Reported, not ranked

Current API price

$2/$6 /M

14,560 pts/$

SWE-Bench Multimodal

1 scored · 0 independently measured · resolve rate

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

59.4

Reported, not ranked

Current API price

$5/$25 /M

3,107 pts/$

Frontier-Bench

1 scored · 0 independently measured · mean reward

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

43.3

Reported, not ranked

Current API price

$5/$25 /M

2,265 pts/$

DeepSearchQA

1 scored · 0 independently measured · F1

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

95

Reported, not ranked

Current API price

$5/$25 /M

4,969 pts/$

HealthBench Professional

1 scored · 0 independently measured · score

Top reported

No independent ranking
Claude Opus 5

Anthropic · Unclassified source · anthropic.com

Score

59.8

Reported, not ranked

Current API price

$5/$25 /M

3,128 pts/$

CyberGym

1 scored · 0 independently measured · score

Top reported

No independent ranking
DeepSeek V4 Flash 0731

DeepSeek · Unclassified source · api-docs.deepseek.com

Score

76.7

Reported, not ranked

Current API price

$0.035/$0.07 /M

938,513 pts/$

DSBench-FullStack

1 scored · 0 independently measured · score

Top reported

No independent ranking
DeepSeek V4 Flash 0731

DeepSeek · Unclassified source · api-docs.deepseek.com

Score

68.7

Reported, not ranked

Current API price

$0.035/$0.07 /M

840,624 pts/$

DSBench-Hard

1 scored · 0 independently measured · score

Top reported

No independent ranking
DeepSeek V4 Flash 0731

DeepSeek · Unclassified source · api-docs.deepseek.com

Score

59.6

Reported, not ranked

Current API price

$0.035/$0.07 /M

729,275 pts/$

GDP.pdf

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Gemini 3.7 Flash

Google DeepMind · Unclassified source · deepmind.google

Score

34

Reported, not ranked

Current API price

$0.75/$3.75 /M

12,141 pts/$

IFEval

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
LongCat-2.0

Meituan · Unclassified source · github.com

Score

90

Reported, not ranked

Current API price

$0.3/$1.20 /M

91,436 pts/$

FORTE

1 scored · 0 independently measured · success rate

Top reported

No independent ranking
LongCat-2.0

Meituan · Unclassified source · github.com

Score

73.2

Reported, not ranked

Current API price

$0.3/$1.20 /M

74,368 pts/$

DeepSearch QA

1 scored · 0 independently measured

Top reported

No independent ranking
Muse Glimmer 30B

Meta · Unclassified source · huggingface.co

Score

74.6

Reported, not ranked

Current API price

$0.2/$0.8 /M

111,343 pts/$

AIME 2026

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Muse Glimmer 30B

Meta · Unclassified source · huggingface.co

Score

94.7

Reported, not ranked

Current API price

$0.2/$0.8 /M

141,343 pts/$

Finance Agent

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Muse Spark 1.1

Meta · Unclassified source · ai.meta.com

Score

57.2

Reported, not ranked

Current API price

$1.25/$4.25 /M

15,198 pts/$

BabyVision

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Muse Spark 1.1

Meta · Unclassified source · ai.meta.com

Score

76.3

Reported, not ranked

Current API price

$1.25/$4.25 /M

20,272 pts/$

MMLU

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Phi-4-mini

Microsoft · Unclassified source · huggingface.co

Score

67.3

Reported, not ranked

Current API price

$0.075/$0.3 /M

251,002 pts/$

τ³-Telecom

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Mistral Medium 3.5

Mistral · Unclassified source · mistral.ai

Score

91.4

Reported, not ranked

Current API price

$1.50/$7.50 /M

16,014 pts/$

Kimi Code Bench

1 scored · 0 independently measured

Top reported

No independent ranking
Kimi K2.7 Code

Moonshot AI · Unclassified source · huggingface.co

Score

62

Reported, not ranked

Current API price

$0.275/$1.10 /M

63,064 pts/$

MCP Mark Verified

1 scored · 0 independently measured · success rate

Top reported

No independent ranking
Kimi K2.7 Code

Moonshot AI · Unclassified source · huggingface.co

Score

81.1

Reported, not ranked

Current API price

$0.275/$1.10 /M

82,492 pts/$

Kimi Claw 24/7 Bench

1 scored · 0 independently measured

Top reported

No independent ranking
Kimi K2.7 Code

Moonshot AI · Unclassified source · huggingface.co

Score

46.9

Reported, not ranked

Current API price

$0.275/$1.10 /M

47,705 pts/$

SpreadsheetBench

1 scored · 0 independently measured · score

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

34.8

Reported, not ranked

Current API price

$2/$8 /M

5,253 pts/$

ZeroBench

1 scored · 0 independently measured · pass@5

Top reported

No independent ranking
Kimi K3

Moonshot AI · Unclassified source · kimi.com

Score

41

Reported, not ranked

Current API price

$2/$8 /M

6,189 pts/$

GPQA

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
Nemotron 3 Ultra 550B A55B

NVIDIA · Unclassified source · huggingface.co

Score

87

Reported, not ranked

Current API price

$0.1/$0.1 /M

419,277 pts/$

FinanceAgent

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-5.4 Pro

OpenAI · Unclassified source · openai.com

Score

61.5

Reported, not ranked

Current API price

$27/$160 /M

502 pts/$

ScreenSpot-Pro

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

92.7

Reported, not ranked

Current API price

$10/$50 /M

2,436 pts/$

BenchCAD

1 scored · 0 independently measured · geometric overlap

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

95.9

Reported, not ranked

Current API price

$10/$50 /M

2,520 pts/$

Terminal-Bench Science

1 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

64.6

Reported, not ranked

Current API price

$10/$50 /M

1,698 pts/$

ExploitBench

1 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

100

Reported, not ranked

Current API price

$10/$50 /M

2,628 pts/$

SRE-Bench

1 scored · 0 independently measured · success rate

Top reported

No independent ranking
GPT-6 Astra

OpenAI · Unclassified source · openai.com

Score

88

Reported, not ranked

Current API price

$10/$50 /M

2,313 pts/$

AIME26

1 scored · 0 independently measured

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

96.67

Reported, not ranked

Current API price

$0.95/$4 /M

29,688 pts/$

LiveCodeBench v6

1 scored · 0 independently measured

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

90.33

Reported, not ranked

Current API price

$0.95/$4 /M

27,740 pts/$

JFBench

1 scored · 0 independently measured

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

37.4

Reported, not ranked

Current API price

$0.95/$4 /M

11,486 pts/$

Translation

1 scored · 0 independently measured

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

52.2

Reported, not ranked

Current API price

$0.95/$4 /M

16,031 pts/$

FairPoliticsQA

1 scored · 0 independently measured

Top reported

No independent ranking
Sakana Namazu

Sakana · Unclassified source · console.sakana.ai

Score

56.3

Reported, not ranked

Current API price

$0.95/$4 /M

17,290 pts/$

ClawEval

1 scored · 0 independently measured · pass^3

Top reported

No independent ranking
Step 3.7 Flash

Stepfun · Unclassified source · static.stepfun.com

Score

67.1

Reported, not ranked

Current API price

$0.185/$1.11 /M

83,668 pts/$

Terminal Bench 2.0

1 scored · 0 independently measured · score

Top reported

No independent ranking
GLM-4.7

Zhipu AI · Unclassified source · huggingface.co

Score

33.4

Reported, not ranked

Current API price

$0.2/$0.8 /M

48,759 pts/$

CritPt

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

20.9

Reported, not ranked

Current API price

$0.3/$1.05 /M

22,656 pts/$

AIME

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

99.2

Reported, not ranked

Current API price

$0.3/$1.05 /M

107,534 pts/$

HMMT

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

94.4

Reported, not ranked

Current API price

$0.3/$1.05 /M

102,331 pts/$

IMOAnswerBench

1 scored · 0 independently measured · accuracy

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

91

Reported, not ranked

Current API price

$0.3/$1.05 /M

98,645 pts/$

PostTrainBench

1 scored · 0 independently measured · score

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

34.3

Reported, not ranked

Current API price

$0.3/$1.05 /M

37,182 pts/$

Tool-Decathlon

1 scored · 0 independently measured · score

Top reported

No independent ranking
GLM-5.2

Zhipu AI · Unclassified source · z.ai

Score

48.2

Reported, not ranked

Current API price

$0.3/$1.05 /M

52,249 pts/$