Skip to content

Benchmark

MTEB(Africa, 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
23
Scores recorded
23
Sources
1
IQ weight
1.59

Score distribution

37.6%
Mean
33.1%
Median
8.2%
Std dev
30.3%
10th percentile
51.4%
90th percentile
29.7%
Min
29.7%55%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.59 for embedding models, measured over 23 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 Multimodal Boards data docs55%
2Multilingual E5 Large InstructClustering· MTEB Multimodal Boards data docs54.3%
3Qwen3 Embedding 4BClustering· MTEB Multimodal Boards data docs51.4%
4E5 Mistral 7B InstructClustering· MTEB Multimodal Boards data docs48.2%
5Qwen3 Embedding 0.6BClustering· MTEB Multimodal Boards data docs47.5%
6Jina Embeddings v5 Omni NanoClustering· MTEB Multimodal Boards data docs47.2%
7Jina Embeddings v3Clustering· MTEB Multimodal Boards data docs38.8%
8Multilingual E5 LargeClustering· MTEB Multimodal Boards data docs37.5%
9BGE M3Clustering· MTEB Multimodal Boards data docs36.8%
10snowflake-arctic-embed-l-v2.0Clustering· MTEB Multimodal Boards data docs36%
11Granite Embedding 278m MultilingualClustering· MTEB Multimodal Boards data docs35.7%
12BGE Large En V1.5Clustering· MTEB Multimodal Boards data docs33.1%
13UAE-Large-V1Clustering· MTEB Multimodal Boards data docs33%
14E5 Base V2Clustering· MTEB Multimodal Boards data docs32%
15BGE Base En V1.5Clustering· MTEB Multimodal Boards data docs31.7%
16bge-small-en-v1.5Clustering· MTEB Multimodal Boards data docs31.7%
17Gte LargeClustering· MTEB Multimodal Boards data docs31.6%
18Gte BaseClustering· MTEB Multimodal Boards data docs31.5%
19nomic-embed-text-v1Clustering· MTEB Multimodal Boards data docs31.4%
20E5 Large V2Clustering· MTEB Multimodal Boards data docs30.5%
21Paraphrase Multilingual MPNet v2Clustering· MTEB Multimodal Boards data docs30.3%
22All MiniLM L6 V2Clustering· MTEB Multimodal Boards data docs29.9%
23nomic-embed-text-v1.5Clustering· MTEB Multimodal Boards data docs29.7%