Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Video Reasoning on VBVR-Bench

97.4Overall Accuracy

Human

24.49643.42362.3581.277Mar 17, 2026Mar 25, 2026Apr 3, 2026Apr 12, 2026Apr 21, 2026Apr 30, 2026May 9, 2026
Updated 22d ago

Evaluation Results

MethodLinks
2026.03
97.4
2026.05
75.7
2026.03
71.6
2026.05
70.7
2026.05
69.4
2026.03
68.5
2026.05
67.1
2026.05
65
2026.03
54.6
2026.03
48
2026.03
40.3
2026.05
40.3
2026.03
37.1
2026.03
36.9
2026.03
31.3
2026.05
30.8
2026.03
27.3
2026.03
27.3