Skip to content

Benchmark

MTEB(nld, 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
25
Scores recorded
25
Sources
1
IQ weight
0.38

Score distribution

69.7%
Mean
66.3%
Median
11.8%
Std dev
57%
10th percentile
86.1%
90th percentile
45.8%
Min
45.8%87%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.38 for embedding models, measured over 25 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard