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

Video Reasoning on VSI-Bench

43.1Accuracy

EvoVid

23.54828.62433.738.776Oct 19, 2025Nov 23, 2025Dec 29, 2025Feb 3, 2026Mar 10, 2026Apr 15, 2026May 21, 2026
Updated 1d ago

Evaluation Results

MethodLinks
2026.05
43.1--
2026.05
42.8--
2026.04
41.9--
2026.05
41.4--
2026.05
40.6--
2026.05
40.1--
2026.05
39.8--
2026.04
39.1--
2026.04
38.922.33-
2026.04
38.9--
2026.04
38.621.77-
2026.04
38.5--
2026.05
38.3--
2026.04
38.1--
2026.05
38--
2026.05
37.4--
2026.04
36.8--
2026.04
35.8--
2025.10
35.6-39.2
2026.05
35.3--
2026.04
34--
2025.10
34--
2025.10
33.8-37
32.4--
31.8--
2026.04
31.8--
31.7--
2026.05
31.7--
2026.05
30.9--
2025.10
30.5-23.7
2025.10
30.3-23.4
30.2--
2026.05
29.5--
2026.05
29.5--
2026.05
29.3--
2025.10
29.2--
28.9--
2025.10
28.5-22.6
2026.05
28.5--
2026.05
28.4--
2025.10
28.1-22.3
2025.10
27.9-21.6
2026.04
27.7--
2026.05
27.7--
2025.10
26.4-21.4
2025.10
26.3-20.4
26.3--
2026.05
26.3--
2026.05
25.1--
2025.10
24.7-17.5
2025.10
24.3-17