Long-context Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
36.8R@10
6
May 8, 2026
59.5R@10
6
May 8, 2026
38Construction Time (s)
6
Mar 5, 2026
51.23F1 Score
6
Feb 26, 2026
52.04F1 Score
6
Feb 26, 2026
53.25F1 Score
6
Feb 26, 2026
51.39F1 Score
6
Feb 26, 2026
29.53F1 Score
6
Feb 26, 2026
84.9Answer F1 Score
6
Feb 26, 2026
85.1Answer F1
6
Feb 26, 2026
0.859Answer F1
6
Feb 26, 2026
66.3EM
6
Feb 26, 2026
74.1CSQA
5
May 21, 2026
73.63Accuracy
5
May 11, 2026
66.7Accuracy
4
Jul 2, 2026
52Accuracy
4
Jul 2, 2026
61.6Accuracy
4
Jul 2, 2026
49.4Accuracy
4
Jul 2, 2026
48Accuracy
4
Jul 2, 2026
50.4Jaccard Score (S-4.5, 3x)
4
Apr 28, 2026
30.2Coursera QA
4
Mar 23, 2026
18.27String Match
3
Jul 8, 2026
11.48String Match
3
Jul 8, 2026
6String Match
3
Jul 8, 2026
20String Match
3
Jul 8, 2026