Math Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
50.4Score
13
Mar 5, 2026
12.3Acc
13
Feb 26, 2026
41.66Pass@1
12
Jun 30, 2026
4.37Speedup
12
Jul 1, 2026
86.67Accuracy (AMC23)
12
Jun 23, 2026
40.4GSM8K Accuracy
12
Jun 18, 2026
96Accuracy (%)
12
May 26, 2026
53.5Mean@32
12
Apr 30, 2026
16.5Mean@32
12
Apr 30, 2026
79Mean Score (k=32)
12
Apr 30, 2026
50Pass@4
12
Apr 28, 2026
68.3Pass@4
12
Apr 28, 2026
65.8Pass@4
12
Apr 28, 2026
43.8Pass@1
12
Apr 7, 2026
79.55Accuracy
12
Mar 9, 2026
43.2Accuracy
12
Mar 9, 2026
40Pass@1
12
Mar 27, 2026
0.813BLEU
12
Feb 26, 2026
5.68Speedup
12
Feb 26, 2026
6.09Speedup
12
Feb 26, 2026
5.15Speedup
12
Feb 26, 2026
53.7Accuracy
12
Feb 26, 2026
97.3Accuracy (AIME 2025)
11
Jul 9, 2026
4.33Brier Score
11
Jun 1, 2026
26.8Worst Case Accuracy (WCA)
11
May 22, 2026