Visual Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
66.54Score
8
May 22, 2026
47.95Overall Accuracy
8
May 13, 2026
29.33Pass@1
8
Apr 7, 2026
73Accuracy
8
Feb 26, 2026
93.9Accuracy
8
Feb 26, 2026
93.27Accuracy
8
Feb 26, 2026
53.8MathVision Accuracy
7
Jun 30, 2026
29.4Solve Rate
7
Jun 12, 2026
79.6Solve Rate
7
Jun 12, 2026
37.8Accuracy
7
May 27, 2026
12Accuracy
7
May 27, 2026
42.3Accuracy
7
May 13, 2026
100Accuracy
7
Feb 26, 2026
92.98Accuracy
7
Feb 26, 2026
99.69Accuracy
7
Feb 27, 2026
97.13Accuracy
7
Feb 27, 2026
71.66ID
7
Feb 26, 2026
68.09Accuracy
7
Feb 26, 2026
83.3Overall Accuracy
6
Jun 24, 2026
83.8Accuracy
6
Jun 16, 2026
61.9VLMsAreBlind Score
6
Apr 3, 2026
15.9Exact Match (EM)
6
Apr 3, 2026
38.8EM
6
Apr 3, 2026
58.6EM
6
Apr 3, 2026
83.8Accuracy
6
Mar 10, 2026