Mathematical Problem Solving
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
42.1Pass@1
8
Apr 21, 2026
54.1Pass@1
8
Apr 21, 2026
26Pass@4
8
Feb 26, 2026
232.4pass@1
8
Feb 26, 2026
0.966Pass@1
8
Feb 26, 2026
80Score
8
Feb 26, 2026
69.01Pass@1 (%)
7
May 8, 2026
12.92Pass@1
7
May 8, 2026
13.96Pass@1 Accuracy
7
May 8, 2026
33.32Pass@1
7
May 8, 2026
42.13Pass@1
7
May 8, 2026
63.9Accuracy
7
Feb 26, 2026
17.6Accuracy
7
Feb 26, 2026
97.6Top-1 Accuracy
6
May 21, 2026
71Accuracy
6
May 13, 2026
43.8Clean Error
6
Apr 15, 2026
70.36Pass@1 Accuracy
6
Apr 1, 2026
65.6MATH Accuracy
6
Feb 26, 2026
88.25Accuracy
6
May 27, 2026
123.3Throughput (tokens/s)
5
Jun 9, 2026
48.3Accuracy
5
Jun 2, 2026
77.3Accuracy (MATH-HARD)
5
Jun 2, 2026
62.2Accuracy
5
Jun 2, 2026
64.44Pass@12
5
May 22, 2026
83.33Pass@12
5
May 22, 2026