Multimodal Math Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
86Accuracy
263
Jun 11, 2026
98.7Accuracy
228
Jun 11, 2026
80.83Accuracy
41
Jun 4, 2026
76.3Score
21
May 11, 2026
65.3Overall Accuracy
16
Feb 26, 2026
69.2ARD
12
Feb 26, 2026
68.58Avg@8 Accuracy
10
Apr 15, 2026
69.53Avg@8 Accuracy
10
Apr 15, 2026
38.9Accuracy
8
Mar 31, 2026
65.3Average Score (DynaMath)
6
May 6, 2026
47.2MathVision Standard Score
6
May 6, 2026
53Accuracy
5
Feb 26, 2026