Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
30.13GPQA-D Score
12
Jun 2, 2026
53Accuracy
12
Jun 2, 2026
0.667Accuracy
12
Jun 2, 2026
73.3Accuracy
12
Jun 2, 2026
78.32Accuracy
12
May 29, 2026
73.9Overall Accuracy
12
May 27, 2026
95Accuracy
12
May 27, 2026
83.3Reasoning Accuracy
12
May 12, 2026
41.17Accuracy (maj@4)
12
Apr 23, 2026
44.4Avg@8
12
Apr 23, 2026
44.5Avg Accuracy@1
12
Apr 23, 2026
73.4Avg@1 (AGI Eval Reasoning)
12
Apr 23, 2026
7.52Speedup
12
Apr 15, 2026
7.23Speedup
12
Apr 15, 2026
7.35Speedup
12
Apr 15, 2026
69.9Accuracy
12
Apr 7, 2026
81.2Accuracy
12
Mar 31, 2026
86.2Accuracy (Wino LOO)
12
Mar 12, 2026
82.4Average Accuracy
12
Mar 12, 2026
88.5Average Accuracy
12
Mar 12, 2026
87.7Average Accuracy
12
Mar 12, 2026
56.9Average Accuracy
12
Mar 12, 2026
94.7Average Accuracy
12
Mar 12, 2026
83.8Accuracy
12
Mar 12, 2026
93.5Accuracy (ARC-e LOO)
12
Mar 12, 2026