Skip to content

Benchmark

MTEB(Multilingual, v2) - Clustering

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

Models measured
33
Scores recorded
33
Sources
1
IQ weight
1.14

Score distribution

45%
Mean
42.2%
Median
7.2%
Std dev
39.8%
10th percentile
53.4%
90th percentile
21.1%
Min
21.1%57.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.14 for embedding models, measured over 33 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 data57.7%
2Qwen3 Embedding 4BClustering· MTEB Leaderboard data57.2%
3Gemini Embedding 001Clustering· MTEB Leaderboard data54.6%
4Jina Embeddings v5 Omni SmallClustering· MTEB Leaderboard data53.4%
5Jina Embeddings v5 Text SmallClustering· MTEB Leaderboard data53.4%
6Jina Embeddings v5 Omni NanoClustering· MTEB Leaderboard data52.7%
7Jina Embeddings v5 Text NanoClustering· MTEB Leaderboard data52.7%
8Qwen3 Embedding 0.6BClustering· MTEB Leaderboard data52.3%
9Multilingual E5 Large InstructClustering· MTEB Leaderboard data50.7%
10E5 Mistral 7B InstructClustering· MTEB Leaderboard data50.6%
11Text Multilingual Embedding 002Clustering· MTEB Leaderboard data47.8%
12embeddinggemma-2Clustering· MTEB Leaderboard data47.6%
13Text Embedding 3 LargeClustering· MTEB Leaderboard data47.1%
14Text Embedding 3 SmallClustering· MTEB Leaderboard data46.1%
15Jina Embeddings v3Clustering· MTEB Leaderboard data45.6%
16snowflake-arctic-embed-l-v2.0Clustering· MTEB Leaderboard data42.8%
17snowflake-arctic-embed-m-v2.0Clustering· MTEB Leaderboard data42.2%
18UAE-Large-V1Clustering· MTEB Leaderboard data41.9%
19Gte BaseClustering· MTEB Leaderboard data41.5%
20BGE Large En V1.5Clustering· MTEB Leaderboard data41.4%
21BGE Base En V1.5Clustering· MTEB Leaderboard data41.4%
22Granite Embedding 278m MultilingualClustering· MTEB Leaderboard data41.4%
23Gte LargeClustering· MTEB Leaderboard data41.4%
24BGE M3Clustering· MTEB Leaderboard data40.9%
25nomic-embed-text-v1Clustering· MTEB Leaderboard data40.8%
26NV-Embed-v2Clustering· MTEB Leaderboard data40.8%
27Paraphrase Multilingual MPNet v2Clustering· MTEB Leaderboard data40.2%
28E5 Base V2Clustering· MTEB Leaderboard data40.1%
29bge-small-en-v1.5Clustering· MTEB Leaderboard data39.9%
30E5 Large V2Clustering· MTEB Leaderboard data39.8%
31nomic-embed-text-v1.5Clustering· MTEB Leaderboard data39.7%
32All MiniLM L6 V2Clustering· MTEB Leaderboard data37.5%
33Jina Embeddings v2 Base ENClustering· MTEB Leaderboard data21.1%