Multimodal Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
31.83Accuracy
8
Jun 15, 2026
81.95Accuracy
8
Jun 15, 2026
69.3Accuracy
8
Jun 15, 2026
19.34Accuracy
8
Jun 15, 2026
2,448Sum Score
8
May 15, 2026
MathVista, WeMath, ChartQA, LogicVista, MMStar, VisPuzzles, RealWorldQA Unseen Degradations 8B model
64.8MathVista Accuracy
8
May 12, 2026
73.3MathVista Score
8
May 12, 2026
78.5MathVista Accuracy
8
May 12, 2026
45.6Overall Accuracy
8
Feb 26, 2026
55.7Accuracy
8
Feb 26, 2026
34.9Accuracy
8
Feb 26, 2026
25.7Accuracy
8
Feb 26, 2026
27.8Accuracy
8
Feb 26, 2026
80.68Accuracy
7
Jun 2, 2026
76.22Accuracy
7
Jun 2, 2026
83.17Attribute Perception Score
7
Mar 4, 2026
41.89MMMU Score
7
Apr 23, 2026
70.1Accuracy
7
Feb 26, 2026
92.1Pass@1
7
Feb 26, 2026
89.5Pass@1
7
Mar 6, 2026
69.8Pass@1
7
Feb 26, 2026
51.2Score
6
Jul 7, 2026
33.9Overall Accuracy
6
Jul 2, 2026
87Pass@1 Score
6
Mar 24, 2026
55.3Accuracy
6
Jun 12, 2026