Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
55Accuracy (GSM8K)
3
Mar 17, 2026
103.63TPS
3
Mar 12, 2026
143.93TPS
3
Mar 12, 2026
79Performance Score
3
Mar 12, 2026
0.032Brier Score
3
Mar 5, 2026
35.4Accuracy (Best-of-64)
3
Mar 5, 2026
58.66Best-of-64 Accuracy
3
Mar 5, 2026
53.66Best-of-64 Accuracy
3
Mar 5, 2026
31.3pass@64
3
Mar 4, 2026
47.8Pass@64
3
Mar 4, 2026
62.1Pass@64
3
Mar 4, 2026
50.1Training Dataset (%)
3
Mar 4, 2026
62.97Win Rate
3
Mar 4, 2026
43.3AIME 2024 Score
3
Mar 4, 2026
0.32Cost ($)
3
Feb 26, 2026
41.9Accuracy
3
Feb 26, 2026
76.4Non-Eng AVG Accuracy (Original)
3
Feb 26, 2026
66Accuracy
3
Feb 26, 2026
81Accuracy
3
Feb 26, 2026
83Accuracy
3
Feb 26, 2026
36F1 Score
3
Feb 26, 2026
93Solve Rate
3
Feb 26, 2026
19.16Avg@32 Accuracy
3
Feb 26, 2026
31.25AIME 2024 Avg@32 Accuracy
3
Feb 26, 2026
0.627Final Eval Reward
3
Feb 26, 2026