Skip to content

Benchmark

BRIGHT-Pro

Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.

Models measured
7
Scores recorded
7
Sources
1
IQ weight
0.54

Score distribution

30.9%
Mean
29.8%
Median
7.2%
Std dev
24.9%
10th percentile
40.8%
90th percentile
20.5%
Min
20.5%41.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.54 for embedding models, measured over 7 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard