Multi-hop Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
69.1Accuracy
8
Jun 17, 2026
79.69Score (7k)
8
Jun 4, 2026
10.8Query Latency (s/q)
8
May 26, 2026
14Query Latency (s/q)
8
May 26, 2026
91.1F1 Score
8
May 26, 2026
73.9F1 Score
8
May 26, 2026
83.9F1 Score
8
May 26, 2026
75.7Best@3
8
May 22, 2026
57.6Task Score
8
May 22, 2026
86.69F1 Score
8
May 14, 2026
42.4Score
8
May 8, 2026
26.933Accuracy
8
Apr 28, 2026
1.1Helmet Correctness Score
8
Apr 21, 2026
1.41Helmet Correctness Score
8
Apr 21, 2026
36Exact Match (HotPotQA)
8
Apr 7, 2026
83Score
8
Mar 13, 2026
48.3F1 Score
8
Feb 26, 2026
803-Hop Accuracy
8
Feb 27, 2026
81.5Overall Accuracy
7
Jun 30, 2026
37.8EM
7
Jun 30, 2026
43.5EM
7
Jun 30, 2026
1Average Rounds
7
Jun 26, 2026
0.14Cost per Pass
7
Jun 23, 2026
0.29Cost
7
Jun 23, 2026
0.09Cost per Pass
7
Jun 23, 2026