Multi-modal Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.68Accuracy
90
May 26, 2026
51.5Score
81
Jul 9, 2026
38.5Accuracy
61
Jun 16, 2026
80.8Accuracy
49
May 26, 2026
72.26Accuracy
48
Apr 21, 2026
47.7Accuracy (%)
45
Mar 24, 2026
85.6Accuracy
36
Apr 23, 2026
63.78Accuracy
28
Apr 21, 2026
73.9Accuracy
28
Mar 4, 2026
92.94Difference Reasoning Accuracy
19
Mar 4, 2026
68.5Traffic Light Accuracy
18
Mar 27, 2026
79.2Accuracy
16
May 11, 2026
86.5Overall Accuracy
14
Jun 15, 2026
72.4Score
13
Jul 1, 2026
75.8Score
13
Jul 1, 2026
52.77GeoQAtest Accuracy
12
Feb 26, 2026
37.9Pass@1
11
May 7, 2026
62.2AV Event Alignment Score
10
Jun 11, 2026
84.7Overall Accuracy
8
Mar 4, 2026
8.9Time
7
Feb 26, 2026
34.1Temporal Reasoning Score
5
Mar 4, 2026
43.7MMStar Score
3
Apr 15, 2026
64SR
3
Feb 26, 2026