Helpfulness evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
89Helpfulness Score
40
Jul 7, 2026
63.2Accuracy
32
May 12, 2026
4.57Average Helpfulness Score
24
Mar 10, 2026
90Helpfulness Score
20
Jul 7, 2026
9.35Helpfulness
18
May 19, 2026
88.8BRR
15
May 8, 2026
73.3Win Rate
14
Jun 16, 2026
70.1Win Rate
14
Jun 16, 2026
86.83BRR Score
12
May 8, 2026
93.1Conversation Score
11
Feb 26, 2026
54.4GPT-Eval Score
10
Apr 3, 2026
77Helpfulness Fraction
7
Feb 26, 2026
84.05Win Rate
5
Mar 5, 2026
80.6Conciseness
3
Jun 2, 2026
90.68HHH Score
3
Feb 26, 2026