Deep Research
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
92.8Mean Correct Rate
8
Feb 26, 2026
97.8Accuracy
8
Feb 26, 2026
92.8Mean Correct Answer Rate
8
Feb 26, 2026
0.92Mean Correct Rate
8
Feb 26, 2026
69.3Overall Score
8
Feb 26, 2026
80Win Rate
8
Feb 26, 2026
3.7Interest Level
7
Feb 26, 2026
40.5Accuracy
6
Feb 26, 2026
1,084Elo Rating
6
Feb 26, 2026
32.33Exact Match (EM)
5
May 28, 2026
4.5Interest Level
5
Feb 26, 2026
40Score
4
Jun 1, 2026
70.1Score
4
Jun 1, 2026
33.02F1 Score
4
Mar 4, 2026
37.69F1 Score
4
Mar 4, 2026
25.07F1 Score
4
Mar 4, 2026