Skip to content

Benchmark

VQAv2 - yes/no

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

96.9%
Mean
96.9%
Median
0%
Std dev
96.9%
10th percentile
96.9%
90th percentile
96.9%
Min

Leaderboard

#ModelScore
1Molmo 72Byes/no· EvalAI data96.9%