Skip to content

Benchmark

MMEB Image - OK-VQA

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
1
Scores recorded
1
Sources
1
IQ weight
No weight published

Score distribution

77.8%
Mean
77.8%
Median
0%
Std dev
77.8%
10th percentile
77.8%
90th percentile
77.8%
Min

Leaderboard

Data credits: MMEB Leaderboard (TIGER-Lab), Apache-2.0 (licence)