Skip to content

Benchmark

MTEB(kor, v1)

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

Models measured
21
Scores recorded
21
Sources
1
IQ weight
1.63

Score distribution

40.4%
Mean
23.9%
Median
24.7%
Std dev
16.2%
10th percentile
70.9%
90th percentile
6.6%
Min
6.6%77.9%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.63 for embedding models, measured over 21 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard