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

Video Captioning on Molmo2 CapTest (test)

56.6F1 Score

GPT-5 mini

5.74418.94732.1545.353Jan 15, 2026Jan 27, 2026Feb 8, 2026Feb 21, 2026Mar 5, 2026Mar 17, 2026Mar 30, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.01
56.6
2026.03
56.6
2026.01
50.1
2026.03
50.1
2026.01
46
2026.03
46
2026.01
43.2
2026.03
43.2
2026.01
42.1
2026.03
42.1
2026.03
40.3
2026.01
40.1
2026.03
40.1
2026.01
39.9
2026.03
39.9
2026.01
36
2026.03
36
2026.01
29.3
2026.03
29.3
2026.01
26.7
2026.03
26.7
2026.01
26
2026.03
26
2026.01
25.4
2026.03
25.4
2026.01
25.2
2026.03
25.2
2026.01
22.8
2026.03
22.8
2026.01
19.9
2026.03
19.9
2026.01
18.4
2026.03
18.4
2026.01
14.8
2026.03
14.8
2026.01
12.3
2026.03
12.3
2026.01
10.9
2026.03
10.9
2026.01
7.8
2026.03
7.8
2026.01
7.7
2026.03
7.7