Mathematical Problem Solving
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
47.1Pass@k
24
May 13, 2026
37.5Pass@k
24
May 13, 2026
79.3Pass@k
24
May 13, 2026
64.7Pass@k
24
May 13, 2026
63Pass@k
24
May 13, 2026
94.27Score
22
Jul 2, 2026
82.5Accuracy (Exact Match)
22
Jun 18, 2026
92MATH-500 Accuracy
22
Jun 16, 2026
85.88MATH500 Score
21
Mar 19, 2026
76Pass@1
21
Apr 1, 2026
59.1Success Rate
20
Feb 26, 2026
0.362Overall Accuracy
20
Feb 26, 2026
42.7Pass@1 Accuracy
19
Jun 30, 2026
89.2Accuracy
18
Jul 7, 2026
47.4Accuracy @ t1
18
Feb 26, 2026
72.9Accuracy
18
Feb 26, 2026
41.7CoT Error
17
Apr 21, 2026
35.4Best-of-N Accuracy
17
Mar 5, 2026
50.98Best-of-N Accuracy
17
Mar 5, 2026
58.66Best-of-N Accuracy
17
Mar 5, 2026
53.66Best-of-N Accuracy
17
Mar 5, 2026
76.07Best-of-N Accuracy
17
Mar 5, 2026
1,046.36AUTC
17
Mar 5, 2026
1,606.4AUTC
17
Mar 5, 2026
1,551.33AUTC Score
17
Mar 5, 2026