Helpfulness
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
90Alpaca Eval (%)
42
Jul 7, 2026
97Average Win Rate
21
Apr 21, 2026
90.71Accuracy
20
Mar 4, 2026
89.7Accuracy
20
Mar 4, 2026
75.15Accuracy
20
Mar 4, 2026
6.64Accuracy
20
Mar 4, 2026
8.75Gemini Score
16
Jun 9, 2026
84.5Win Rate
6
Mar 5, 2026
91.6Win Rate
5
Mar 5, 2026
91.62Win Rate
5
Mar 5, 2026
2.542Reward
4
Feb 26, 2026