Skip to content

Benchmark

VQAv2 - number

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

78.5%
Mean
78.5%
Median
0%
Std dev
78.5%
10th percentile
78.5%
90th percentile
78.5%
Min

Leaderboard

#ModelScore
1Molmo 72Bnumber· EvalAI data78.5%