Skip to content

Benchmark

ViDoRe(v3.1) - Private split

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

Models measured
9
Scores recorded
9
Sources
1
IQ weight
1.37

Score distribution

49.6%
Mean
53.8%
Median
9.5%
Std dev
44.1%
10th percentile
56.7%
90th percentile
26.5%
Min
26.5%60%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.37 for embedding models, measured over 9 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard