GPT-4o mini Benchmark Details

GPT-4o mini currently shows benchmark results led by MBPP (4 / 69, score 87.20), HumanEval (19 / 101, score 87.20), GSM8K (15 / 68, score 91.30).

Benchmark Results

GPT-4o mini

Benchmark Results

Knowledge Exams

3 evaluations

Benchmark / mode

Score

Rank/total

MMLU

Standard Mode

48 / 124

MMLU-Pro

Standard Mode

61.70

120 / 175

MMLU-Pro

unknown

63.09

117 / 175

Mathematics

2 evaluations

Benchmark / mode

Score

Rank/total

GSM8K

Standard Mode

91.30

15 / 68

MATH

Standard Mode

70.20

19 / 42

Code Generation & Editing

2 evaluations

Benchmark / mode

Score

Rank/total

HumanEval

Standard Mode

87.20

19 / 101

MBPP

Standard Mode

87.20

4 / 69

Scientific Reasoning

2 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Standard Mode

41.10

240 / 253

GPQA

Standard Mode

40.20

14 / 17

Honesty & Factuality

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleQA

Standard Mode

9.50

43 / 46

Writing

1 evaluations

Benchmark / mode

Score

Rank/total

Creative Writing

Standard Mode

872.80

96 / 110

Visual Understanding

5 evaluations

Benchmark / mode

Score

Rank/total

GeoBench ACW

Standard Mode

15 / 20

MMMU

unknown

59.40

61 / 74

MMMU-Pro

Standard Mode

41.50

221 / 229

MMMU-Pro

unknown

37.60

225 / 229

VPCT

Standard Mode

23 / 24

Commonsense

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Standard Mode

10.70

96 / 96

Agentic Development

1 evaluations

Benchmark / mode

Score

Rank/total

Aider-Polyglot

Standard Mode

3.60

59 / 59

Tool Orchestration

1 evaluations

Benchmark / mode

Score

Rank/total

Pinch Bench

Thinking EnabledTools

28 / 38

Service Workflows

1 evaluations

Benchmark / mode

Score

Rank/total

τ³-Banking

Standard ModeTools

2.90

166 / 167

ML Engineering

1 evaluations

Benchmark / mode

Score

Rank/total

WeirdML v2

Standard ModeTools

11.76

51 / 52

Logical Planning

1 evaluations

Benchmark / mode

Score

Rank/total

BALROG

Standard ModeTools

17.40

12 / 12

Capability Indices

1 evaluations

Benchmark / mode

Score

Rank/total

ECI

unknown

126.56

123 / 167

Compare with other models