Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
83Target Accuracy
20
Jun 23, 2026
51.3PolyMath Score
20
Jun 5, 2026
95.1ARC-E Accuracy
20
Jun 5, 2026
50PolyMath Score
20
Jun 5, 2026
85.7ARC-E Accuracy
20
Jun 5, 2026
58.8Accuracy
20
Jun 4, 2026
100Head-to-head Win %
20
May 21, 2026
78.75Avg@k Score
20
Apr 20, 2026
75.42Avg@k
20
Apr 20, 2026
79.17Average @k Accuracy
20
Apr 20, 2026
34OpenBookQA Accuracy
20
Feb 26, 2026
60.3Average Accuracy
20
Feb 26, 2026
613Trace Count
20
Feb 26, 2026
100Consistency
20
Feb 26, 2026
681Trace Count
20
Feb 26, 2026
43.55Overall Average Score
20
Feb 26, 2026
53.53Exact Match (EM)
20
Feb 26, 2026
69.85Accuracy
20
Jul 2, 2026
68.6Accuracy
20
Feb 26, 2026
34.43Score
20
Feb 26, 2026
49.58AIME 2025 Accuracy
19
May 1, 2026
57F1 Score
19
Feb 26, 2026
100Accuracy
19
May 13, 2026
61.6Math Average Score
18
May 29, 2026
70.3Accuracy
18
Apr 15, 2026