Skip to content

Benchmark

ChemTEB(v1.1) - Clustering

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

Models measured
27
Scores recorded
27
Sources
1
IQ weight
0.33

Score distribution

45.7%
Mean
47.8%
Median
7%
Std dev
34.3%
10th percentile
52.2%
90th percentile
32.9%
Min
32.9%59.9%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.33 for embedding models, measured over 27 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Text Embedding 3 LargeClustering· MTEB Extra Text Boards data59.9%
2Embed English V3.0Clustering· MTEB Extra Text Boards data55.7%
3Embed Multilingual V3.0Clustering· MTEB Extra Text Boards data53.1%
4Jina Embeddings v5 Text SmallClustering· MTEB Extra Text Boards data52.2%
5Text Embedding Ada 002Clustering· MTEB Extra Text Boards data50.8%
6Qwen3 Embedding 0.6BClustering· MTEB Extra Text Boards data50.7%
7All Mpnet Base V2Clustering· MTEB Extra Text Boards data50.5%
8Qwen3 Embedding 8BClustering· MTEB Extra Text Boards data50.1%
9All MiniLM L12 V2Clustering· MTEB Extra Text Boards data49.7%
10nomic-embed-text-v1.5Clustering· MTEB Extra Text Boards data49.6%
11Titan Text Embeddings V2Clustering· MTEB Extra Text Boards data49.3%
12Text Embedding 3 SmallClustering· MTEB Extra Text Boards data48.9%
13All MiniLM L6 V2Clustering· MTEB Extra Text Boards data48.4%
14nomic-embed-text-v1Clustering· MTEB Extra Text Boards data47.8%
15BGE Base En V1.5Clustering· MTEB Extra Text Boards data46.6%
16bge-large-enClustering· MTEB Extra Text Boards data44.9%
17Titan Embeddings G1 - TextClustering· MTEB Extra Text Boards data44.8%
18BGE M3Clustering· MTEB Extra Text Boards data44.7%
19Jina Embeddings v5 Omni NanoClustering· MTEB Extra Text Boards data43.1%
20Multi Qa Mpnet Base Dot V1Clustering· MTEB Extra Text Boards data42.1%
21Qwen3 Embedding 4BClustering· MTEB Extra Text Boards data41.2%
22BGE Large En V1.5Clustering· MTEB Extra Text Boards data39.2%
23E5 Large V2Clustering· MTEB Extra Text Boards data38.2%
24bge-small-en-v1.5Clustering· MTEB Extra Text Boards data34.3%
25Multilingual E5 Large InstructClustering· MTEB Extra Text Boards data33.4%
26E5 Base V2Clustering· MTEB Extra Text Boards data33%
27Multilingual E5 LargeClustering· MTEB Extra Text Boards data32.9%