Skip to content

Benchmark

MTEB(deu, v1) - Clustering

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

Models measured
29
Scores recorded
29
Sources
1
IQ weight
0.77

Score distribution

29.1%
Mean
27.6%
Median
9.2%
Std dev
18.5%
10th percentile
41%
90th percentile
7.1%
Min
7.1%43.4%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.77 for embedding models, measured over 28 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen3 Embedding 8BClustering· MTEB Leaderboard data43.4%
2Qwen3 Embedding 4BClustering· MTEB Leaderboard data43.1%
3Jina Embeddings v5 Text SmallClustering· MTEB Leaderboard data41.3%
4Jina Embeddings v5 Text NanoClustering· MTEB Leaderboard data41%
5E5 Mistral 7B InstructClustering· MTEB Leaderboard data40.9%
6Multilingual E5 Large InstructClustering· MTEB Leaderboard data40.2%
7Qwen3 Embedding 0.6BClustering· MTEB Leaderboard data37.4%
8Jina Embeddings v3Clustering· MTEB Leaderboard data36.8%
9Multilingual E5 LargeClustering· MTEB Leaderboard data33.5%
10snowflake-arctic-embed-l-v2.0Clustering· MTEB Leaderboard data32.8%
11snowflake-arctic-embed-m-v2.0Clustering· MTEB Leaderboard data31.6%
12Granite Embedding 278m MultilingualClustering· MTEB Leaderboard data31.1%
13E5 Large V2Clustering· MTEB Leaderboard data28.3%
14Gte LargeClustering· MTEB Leaderboard data27.8%
15BGE Large En V1.5Clustering· MTEB Leaderboard data27.6%
16UAE-Large-V1Clustering· MTEB Leaderboard data27.3%
17E5 Base V2Clustering· MTEB Leaderboard data27.1%
18Paraphrase Multilingual MPNet v2Clustering· MTEB Leaderboard data26%
19nomic-embed-text-v1Clustering· MTEB Leaderboard data25.6%
20Gte BaseClustering· MTEB Leaderboard data25.5%
21Jina Embedding B EN v1Clustering· MTEB Leaderboard data25.3%
22BGE Base En V1.5Clustering· MTEB Leaderboard data25.2%
23nomic-embed-text-v1.5Clustering· MTEB Leaderboard data24.9%
24bge-small-en-v1.5Clustering· MTEB Leaderboard data24.1%
25All Mpnet Base V2Clustering· MTEB Leaderboard data23.8%
26All MiniLM L12 V2Clustering· MTEB Leaderboard data18.5%
27All MiniLM L6 V2Clustering· MTEB Leaderboard data18.2%
28Text2vec Base ChineseClustering· MTEB Leaderboard data7.6%
29Nemotron 3 Embed 1bClustering· MTEB Leaderboard data7.1%