Skip to content

Benchmark explainers

What each benchmark actually measures

The benchmarks everyone argues about, explained from the record — what they measure, how they're scored, and how much to trust them.

integrity grade B

What is SWE-bench Verified?

32 models scored · top score 83.47

integrity grade C

What is MMLU?

102 models scored · top score 84.13

integrity grade B

What is GPQA diamond?

194 models scored · top score 95.2

integrity grade A

What is ARC-AGI-2?

84 models scored · top score 95

integrity grade B

What is ARC-AGI?

85 models scored · top score 98.5

integrity grade A

What is AIME?

8 models scored · top score 79.8

integrity grade A

What is FrontierMath Tier 4?

65 models scored · top score 100

integrity grade A

What is SimpleQA Verified?

81 models scored · top score 76

integrity grade A

What is Chess Puzzles?

135 models scored · top score 72.64

integrity grade A

What is LiveCodeBench?

8 models scored · top score 65.9

integrity grade A

What is HLE?

42 models scored · top score 52.52

integrity grade C

What is GSM8K?

56 models scored · top score 92

integrity grade A

What is APEX-Agents?

49 models scored · top score 82.4

integrity grade B

What is DeepResearch Bench?

24 models scored · top score 55.3

integrity grade B

What is Cybench?

19 models scored · top score 93

integrity grade B

What is Lech Mazur Writing?

43 models scored · top score 86

integrity grade B

What is OTIS Mock AIME 2024-2025?

180 models scored · top score 100

integrity grade B

What is MATH level 5?

85 models scored · top score 98.13

integrity grade A

What is DTBench?

154 models scored · top score 98.22

integrity grade A

What is LMCA?

123 models scored · top score 80.27

integrity grade A

What is WeirdML?

119 models scored · top score 93.57

integrity grade A

What is Aider polyglot?

45 models scored · top score 88

integrity grade B

What is Terminal Bench?

35 models scored · top score 84.72

integrity grade B

What is METR Time Horizons?

34 models scored · top score 78.86

integrity grade A

What is FrontierMath Tiers 1–3?

84 models scored · top score 93.68

integrity grade A

What is SimpleBench?

80 models scored · top score 78.28

integrity grade B

What is ProofBench?

69 models scored · top score 100

integrity grade B

What is BBH?

41 models scored · top score 85.6