Skip to content

Benchmark

JMTEB(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
29
Scores recorded
29
Sources
1
IQ weight
0.70

Score distribution

36.7%
Mean
42.6%
Median
15.5%
Std dev
18.2%
10th percentile
53.4%
90th percentile
12.7%
Min
12.7%57%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.70 for embedding models, measured over 29 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1E5 Mistral 7B InstructClustering· MTEB Extra Text Boards data57%
2Qwen3 Embedding 8BClustering· MTEB Extra Text Boards data56.1%
3Multilingual E5 Large InstructClustering· MTEB Extra Text Boards data54.4%
4Qwen3 Embedding 4BClustering· MTEB Extra Text Boards data53.4%
5Qwen3 Embedding 0.6BClustering· MTEB Extra Text Boards data52.1%
6Jina Embeddings v5 Omni NanoClustering· MTEB Extra Text Boards data52%
7Jina Embeddings v5 Omni SmallClustering· MTEB Extra Text Boards data51.9%
8Jina Embeddings v5 Text SmallClustering· MTEB Extra Text Boards data51.8%
9Text Embedding 3 LargeClustering· MTEB Extra Text Boards data50.4%
10Text Embedding 3 SmallClustering· MTEB Extra Text Boards data50.1%
11Granite Embedding 278m MultilingualClustering· MTEB Extra Text Boards data48.8%
12Jina Embeddings v3Clustering· MTEB Extra Text Boards data47.7%
13snowflake-arctic-embed-m-v2.0Clustering· MTEB Extra Text Boards data43.9%
14snowflake-arctic-embed-l-v2.0Clustering· MTEB Extra Text Boards data43.4%
15BGE M3Clustering· MTEB Extra Text Boards data42.6%
16Paraphrase Multilingual MPNet v2Clustering· MTEB Extra Text Boards data42.4%
17NV-Embed-v2Clustering· MTEB Extra Text Boards data38.6%
18nomic-embed-text-v1Clustering· MTEB Extra Text Boards data23.1%
19E5 Large V2Clustering· MTEB Extra Text Boards data22.9%
20nomic-embed-text-v1.5Clustering· MTEB Extra Text Boards data21.6%
21Gte BaseClustering· MTEB Extra Text Boards data20%
22E5 Base V2Clustering· MTEB Extra Text Boards data19.9%
23Gte LargeClustering· MTEB Extra Text Boards data19.8%
24BGE Large En V1.5Clustering· MTEB Extra Text Boards data18.5%
25UAE-Large-V1Clustering· MTEB Extra Text Boards data18.4%
26BGE Base En V1.5Clustering· MTEB Extra Text Boards data18.2%
27bge-small-en-v1.5Clustering· MTEB Extra Text Boards data17.6%
28All MiniLM L6 V2Clustering· MTEB Extra Text Boards data16.2%
29Jina Embeddings v2 Base ENClustering· MTEB Extra Text Boards data12.7%