Retrieval-Augmented Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
76.6Average Performance Score
55
Apr 24, 2026
51.8Reliability Score (RS)
52
Jun 4, 2026
90Accuracy (Clean)
45
May 20, 2026
74.02Accuracy
42
May 22, 2026
100NQ Score
42
Feb 26, 2026
87NQ Score
37
Feb 26, 2026
65.04F1 Score
28
Jun 4, 2026
77.1Accuracy
23
Feb 26, 2026
5.48CR
21
Apr 15, 2026
65Win Rate
18
Feb 26, 2026
76.2Win Rate
18
Feb 26, 2026
95Win Rate
18
Feb 26, 2026
85.6Win Rate
18
Feb 26, 2026
74Overall Score
18
Feb 26, 2026
53.7RAG Score (8k Context)
16
Mar 19, 2026
58.2Win Rate
16
Feb 26, 2026
86.51Language Consistency
15
Jul 7, 2026
96.45Language Consistency
13
Jul 7, 2026
65.4Performance at 8k Context Length
13
Mar 19, 2026
78.02Recall
12
Mar 13, 2026
96.7Comprehensiveness
12
Feb 26, 2026
95.9Comprehensiveness
12
Feb 26, 2026
97.6Comprehensiveness
12
Feb 26, 2026
66.3Accuracy
12
Feb 26, 2026
45.2Accuracy
12
Feb 26, 2026