Skip to content

Benchmark

ViDoRe(v3) - Private split

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

Models measured
6
Scores recorded
6
Sources
1
IQ weight
1.10

Score distribution

52.6%
Mean
55.7%
Median
5.6%
Std dev
45.2%
10th percentile
59.6%
90th percentile
45%
Min
45%59.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.10 for embedding models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard