GPT-4.5 Benchmark Details

GPT-4.5 currently shows benchmark results led by MMLU-Pro (21 / 176, score 86.10), SimpleQA (9 / 47, score 62.50), MMMU (34 / 74, score 74.40).

Benchmark Results

GPT-4.5

Benchmark Results

General Knowledge

2 evaluations

Benchmark / mode

Score

Rank/total

MMLU-Pro

Standard Mode

86.10

21 / 176

HLE

Standard Mode

5.44

471 / 565

Other

1 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Standard Mode

71.40

309 / 463

Common Sense

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleQA

Standard Mode

62.50

9 / 47

Coding and Software Engineer

3 evaluations

Benchmark / mode

Score

Rank/total

LiveCodeBench

Standard Mode

46.40

179 / 251

SWE-bench Verified

Standard Mode

108 / 116

IC SWE-Lancer (Diamond)

Standard Mode

32.60

5 / 8

Math and Reasoning

2 evaluations

Benchmark / mode

Score

Rank/total

MATH-500

Standard Mode

90.70

36 / 46

AIME 2024

Standard Mode

36.70

53 / 62

Writing and Creative Capabilities

1 evaluations

Benchmark / mode

Score

Rank/total

Creative Writing

Standard Mode

1255.30

77 / 106

Multimodal Understanding

1 evaluations

Benchmark / mode

Score

Rank/total

MMMU

unknown

74.40

34 / 74

Common Sense Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Standard Mode

34.50

73 / 93

Agent Level Benchmark

1 evaluations

Benchmark / mode

Score

Rank/total

Aider-Polyglot

Standard Mode

44.90

37 / 59

Compare with other models