Reading Comprehension
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
58.31Accuracy
4
Feb 26, 2026
48.01Accuracy
4
Feb 26, 2026
63.9Accuracy
3
May 25, 2026
85.3Accuracy
3
May 12, 2026
18.95Throughput (TPS)
3
Apr 21, 2026
72.57BoolQ Accuracy
3
Mar 18, 2026
64.22Accuracy
3
Mar 6, 2026
83.1EM
3
Feb 26, 2026
87.6Accuracy
3
Feb 26, 2026
58.92Exact Match
3
Feb 26, 2026
347.3CIDEr
3
Feb 26, 2026
81.6EM
3
Feb 26, 2026
62.54EM
3
Feb 26, 2026
39.7EM
3
Feb 26, 2026
68.88Accuracy
3
Feb 26, 2026
73.1Accuracy
3
Feb 26, 2026
72.3Accuracy
3
Feb 26, 2026
42.5Performance (32k Context)
2
Jun 9, 2026
73Score
2
May 15, 2026
30Accuracy
2
Mar 31, 2026
50.36Score
2
Mar 12, 2026
18.5F1 Score
2
Feb 26, 2026
98.01Accuracy
2
Feb 26, 2026
98.03Accuracy
2
Feb 26, 2026
59.3Overall Score
2
Feb 26, 2026