OpenAI o3-mini Benchmark Details

OpenAI o3-mini currently shows benchmark results led by MMLU (43 / 124, score 84.90), Aider-Polyglot (21 / 59, score 60.40), AIME2025 (53 / 110, score 86.50).

Benchmark Results

OpenAI o3-mini

Benchmark Results

Knowledge Exams

2 evaluations

Benchmark / mode

Score

Rank/total

MMLU

Thinking Enabled

84.90

43 / 124

HLE

Thinking Enabled

13.40

184 / 235

Abstract Generalization

6 evaluations

Benchmark / mode

Score

Rank/total

ARC-AGI-1

Thinking Level · Low

14.50

180 / 193

ARC-AGI-1

Thinking Level · Medium

22.33

171 / 193

ARC-AGI-1

Thinking Level · High

34.50

160 / 193

ARC-AGI-2

Thinking Level · Low

176 / 181

ARC-AGI-2

Thinking Level · Medium

2.08

156 / 181

ARC-AGI-2

Thinking Level · High

152 / 181

Scientific Reasoning

2 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Thinking Enabled

70.60

179 / 253

CritPt

Thinking Level · High

0.30

186 / 204

Repository Engineering

1 evaluations

Benchmark / mode

Score

Rank/total

SWE-bench Verified

Thinking Enabled

40.80

106 / 116

Mathematics

6 evaluations

Benchmark / mode

Score

Rank/total

MATH-500

Thinking Enabled

95.80

25 / 46

AIME2025

Thinking Enabled

86.50

53 / 110

AIME 2024

Thinking Enabled

42 / 61

FrontierMath v2

Thinking Level · High

18.60

52 / 58

FrontierMath - Tier 4

Thinking Level · High

4.20

40 / 80

FrontierMath Tier 4 v2

Thinking Level · High

41 / 42

Algorithmic Coding

1 evaluations

Benchmark / mode

Score

Rank/total

CodeForces

Thinking Enabled

2073

15 / 21

Commonsense

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Thinking Level · High

22.80

87 / 96

Agentic Development

4 evaluations

Benchmark / mode

Score

Rank/total

Aider-Polyglot

Thinking Level · Medium

53.80

29 / 59

Aider-Polyglot

Thinking Level · High

60.40

21 / 59

Terminal Bench Hard

Thinking EnabledTools

6.80

194 / 244

Terminal Bench Hard

Thinking Level · HighTools

6.10

204 / 244

Scientific Computing

1 evaluations

Benchmark / mode

Score

Rank/total

SciCode

Thinking Level · High

42.80

100 / 134

Service Workflows

1 evaluations

Benchmark / mode

Score

Rank/total

τ³-Banking

Thinking Level · HighTools

5.20

157 / 167

Documents & Charts

1 evaluations

Benchmark / mode

Score

Rank/total

GDP.pdf

Thinking Level · High

110 / 122

Capability Indices

1 evaluations

Benchmark / mode

Score

Rank/total

ECI

unknown

140.35

93 / 167

Compare with other models