Skip to content

Benchmark

MTEB(Code, v1) - Mean (TaskType)

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
1.46

Score distribution

71.1%
Mean
75.4%
Median
9.6%
Std dev
55.8%
10th percentile
80.1%
90th percentile
48.4%
Min
48.4%80.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.46 for embedding models, measured over 13 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard