Science Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
95.1Accuracy
243
Apr 15, 2026
59.22Accuracy (GPQA)
72
May 29, 2026
83Accuracy
72
Jun 17, 2026
97Accuracy
65
Jun 16, 2026
91.9Pass@1
57
Jun 11, 2026
64.44Accuracy
56
May 14, 2026
76.3Accuracy
52
May 26, 2026
70.2Pass@1
50
Apr 10, 2026
54.67Accuracy (GPQA)
43
Jul 2, 2026
58.9Accuracy
27
Apr 14, 2026
69.4GPQA Score
27
Apr 16, 2026
72.5Accuracy
25
May 15, 2026
89.6Accuracy
25
Jul 2, 2026
30.9r* Accuracy
24
May 7, 2026
69.2Accuracy
18
Jun 9, 2026
81.4Score
17
Mar 4, 2026
44.8Avg@4 Accuracy
17
Feb 26, 2026
39.4Avg@4
16
Feb 26, 2026
84.04Accuracy
15
Feb 26, 2026
54.6GPQA Diamond
13
May 18, 2026
66.49Accuracy
12
Jun 30, 2026
6,858Response Length (Tokens)
12
Jun 17, 2026
1,956Response Length (Tokens)
12
Jun 17, 2026
59.59Pass@1
12
Feb 26, 2026
71.6AUCOAA
11
Feb 26, 2026