Skip to content

Benchmark

MTEB(Multilingual, v2) - STS

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

Models measured
26
Scores recorded
26
Sources
1
IQ weight
0.92

Score distribution

69.9%
Mean
74%
Median
8.8%
Std dev
59.5%
10th percentile
79.4%
90th percentile
56.1%
Min
56.1%81.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.92 for embedding models, measured over 26 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard