Skip to content

Benchmark

MTEB(cmn, v1) - STS

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

Models measured
13
Scores recorded
13
Sources
1
IQ weight
1.33

Score distribution

52%
Mean
54.5%
Median
6.6%
Std dev
41.7%
10th percentile
61.3%
90th percentile
40.6%
Min
40.6%63.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.33 for embedding models, measured over 12 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard