Vision-Language Modeling
Benchmarks
DatasetSOTA methodMetricTrendResultsLast updated
3.8Perplexity
28
Feb 26, 2026
5.1PPL
28
Feb 26, 2026
82.6Accuracy
16
Jun 17, 2026
60.5Accuracy
16
Jun 17, 2026
42Accuracy
16
Jun 17, 2026
26Accuracy
16
Jun 17, 2026
48.8Accuracy
16
Jun 17, 2026
53.2Accuracy
16
Jun 17, 2026
50.5Accuracy
16
Jun 17, 2026
76Accuracy
16
Jun 17, 2026
84.6Accuracy
16
Jun 17, 2026
79.5Accuracy
16
Jun 17, 2026
52.7Accuracy
16
Jun 17, 2026
73.9Accuracy
16
Jun 17, 2026
18.6Accuracy
16
Jun 17, 2026
85.3Accuracy
16
Jun 17, 2026
75.5PASS@1 Accuracy
9
Jun 17, 2026
4.51PPL
7
Feb 26, 2026
20.1Avg Relative Improvement
3
Feb 26, 2026