Visual Understanding
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
76.9MM-Vet Score
190
Jul 7, 2026
2,321MME Score
54
Mar 23, 2026
75.1SEED Score
36
Jun 23, 2026
69.86Accuracy
28
Jun 30, 2026
100Avg Relative Accuracy
26
Apr 15, 2026
1,687MME Perception Score
24
Feb 26, 2026
1,699MME-P Score
18
Mar 31, 2026
0.942Avg@8 EM
18
Feb 26, 2026
86.6Avg@8 Exact Match
17
Feb 26, 2026
90.2Avg@8 Exact Match
17
Feb 26, 2026
65.9Accuracy (Clean)
16
May 22, 2026
1,638MME-P Score
15
Jun 23, 2026
86.96Accuracy
15
May 12, 2026
90.67Jigsaw Accuracy
14
Apr 15, 2026
79Accuracy
14
May 12, 2026
85.9Average Score
13
Feb 26, 2026
72.7Pass@1 Exact Match
13
Feb 26, 2026
73.3Accuracy
11
Mar 4, 2026
80.67Accuracy
11
Mar 4, 2026
85.86Accuracy
11
Mar 4, 2026
75.83Accuracy
11
Mar 4, 2026
76.7Accuracy
10
Feb 26, 2026
85Score
9
Feb 26, 2026
65.29MCQ (low)
8
Feb 26, 2026
1,075.5Score
8
Feb 26, 2026