Video Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
55Average Score
9
May 12, 2026
68.5Accuracy
9
Apr 21, 2026
40.52Accuracy (Long)
8
May 28, 2026
31.8Accuracy
8
Feb 26, 2026
37.8Accuracy
8
Feb 26, 2026
33.8Accuracy
8
Feb 26, 2026
55.4Accuracy
7
Jul 2, 2026
77.8Accuracy
7
Jul 2, 2026
51Accuracy
7
Apr 21, 2026
5.2Runtime (s)
7
Apr 16, 2026
40.4Accuracy
7
Apr 14, 2026
60.4Accuracy
7
Mar 5, 2026
66.5Accuracy
6
Jul 2, 2026
58.7Accuracy
6
May 26, 2026
9.4Non-Linear Score
6
May 8, 2026
65.8Accuracy
6
May 8, 2026
67.2Accuracy
5
Apr 14, 2026
100Body Accuracy
5
Feb 26, 2026
36.3Overall Accuracy
4
Apr 16, 2026
78.5Accuracy
4
Apr 14, 2026
2.14Throughput (it/s)
3
May 26, 2026
27.1F1 Score
2
Feb 26, 2026