Math
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
96.06Accuracy
16
Jun 2, 2026
69.2Avg@16
16
Jun 9, 2026
76.5Avg@16
16
Jun 9, 2026
100Pass@16
16
Mar 6, 2026
67.7Accuracy
15
May 15, 2026
80Accuracy
15
May 15, 2026
92.6GSM8K Accuracy
15
Mar 20, 2026
8.5Speedup
14
Jul 9, 2026
8.08Speedup
14
Jul 9, 2026
94.2Accuracy (%)
14
May 27, 2026
92.04Accuracy
13
May 21, 2026
97.26Accuracy
13
May 21, 2026
44Accuracy
13
Feb 27, 2026
17.3Avg@16
12
Jun 9, 2026
42.8Avg@16
12
Jun 9, 2026
44.5Avg@16
12
Jun 9, 2026
20Score
12
Jun 2, 2026
74.2Accuracy (MATH500)
12
Jun 2, 2026
62.5Accuracy
12
Jun 2, 2026
7.96Speedup
12
May 29, 2026
100True Workspace Rate
12
May 25, 2026
49.1Accuracy
11
May 12, 2026
59.87Solve Rate
11
Feb 26, 2026
94.97Solve Rate
11
Feb 26, 2026
72.5Accuracy (AMC 23)
10
Jun 2, 2026