Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
41.6Accuracy
59
May 18, 2026
44.7Accuracy
59
Mar 13, 2026
86.7Score
58
May 29, 2026
96.23Accuracy
58
May 20, 2026
66.2Accuracy
58
Jun 5, 2026
67Accuracy
58
Jun 19, 2026
60.44Accuracy
57
Jun 9, 2026
54.9Accuracy (@avg1)
57
Mar 5, 2026
96.9Accuracy
57
May 13, 2026
57.02Accuracy
57
May 22, 2026
92.1Accuracy
57
Feb 26, 2026
89.15Accuracy
57
Feb 27, 2026
71.3Accuracy
57
Feb 26, 2026
72.28Accuracy
57
Feb 26, 2026
90.12Accuracy
57
Feb 26, 2026
92.37Accuracy
57
Feb 26, 2026
65.81Accuracy
57
Mar 4, 2026
58.2Accuracy
56
May 27, 2026
93Accuracy (%)
56
Jun 9, 2026
92Accuracy
55
Jul 7, 2026
74.8MathVista
55
May 20, 2026
92.8Accuracy
55
Jun 18, 2026
0.992Mean@1
55
Mar 6, 2026
Mathematical Reasoning Suite (AMC, AIME 2024, AIME 2025, Minerva, MATH, Olympiad) various (test val)
62.1Average Score
55
Feb 26, 2026
719Pass@1
54
Jul 7, 2026