Skip to content

Benchmark

MTEB(Africa, v1) - BitextMining

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

Models measured
36
Scores recorded
36
Sources
1
IQ weight
0.66

Score distribution

32.3%
Mean
31%
Median
20.2%
Std dev
13.5%
10th percentile
62%
90th percentile
0.33%
Min
0.33%85%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.66 for embedding models, measured over 36 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Multilingual E5 Large InstructBitextMining· MTEB Multimodal Boards data docs85%
2Multilingual E5 LargeBitextMining· MTEB Multimodal Boards data docs78.9%
3Gemini Embedding 001BitextMining· MTEB Multimodal Boards data docs71.8%
4Qwen3 Embedding 8BBitextMining· MTEB Multimodal Boards data docs62%
5Text Multilingual Embedding 002BitextMining· MTEB Multimodal Boards data docs58.1%
6Qwen3 Embedding 4BBitextMining· MTEB Multimodal Boards data docs57.4%
7Qwen3 Embedding 0.6BBitextMining· MTEB Multimodal Boards data docs49.5%
8E5 Mistral 7B InstructBitextMining· MTEB Multimodal Boards data docs46.2%
9Jina Embeddings v5 Omni SmallBitextMining· MTEB Multimodal Boards data docs39.2%
10Jina Embeddings v5 Text SmallBitextMining· MTEB Multimodal Boards data docs39.2%
11Text Embedding 3 LargeBitextMining· MTEB Multimodal Boards data docs38.8%
12Jina Embeddings v3BitextMining· MTEB Multimodal Boards data docs37%
13Jina Embeddings v5 Omni NanoBitextMining· MTEB Multimodal Boards data docs37%
14Jina Embeddings v5 Text NanoBitextMining· MTEB Multimodal Boards data docs37%
15snowflake-arctic-embed-l-v2.0BitextMining· MTEB Multimodal Boards data docs36.9%
16embeddinggemma-2BitextMining· MTEB Multimodal Boards data docs36.7%
17Text Embedding 3 SmallBitextMining· MTEB Multimodal Boards data docs32%
18NV-Embed-v2BitextMining· MTEB Multimodal Boards data docs31%
19Granite Embedding 278m MultilingualBitextMining· MTEB Multimodal Boards data docs30.3%
20snowflake-arctic-embed-m-v2.0BitextMining· MTEB Multimodal Boards data docs22.1%
21Paraphrase Multilingual MPNet v2BitextMining· MTEB Multimodal Boards data docs21.4%
22E5 Large V2BitextMining· MTEB Multimodal Boards data docs21.1%
23E5 Base V2BitextMining· MTEB Multimodal Boards data docs19.4%
24nomic-embed-text-v1BitextMining· MTEB Multimodal Boards data docs16.4%
25Gte LargeBitextMining· MTEB Multimodal Boards data docs16.3%
26UAE-Large-V1BitextMining· MTEB Multimodal Boards data docs16.2%
27BGE Large En V1.5BitextMining· MTEB Multimodal Boards data docs16.1%
28Gte BaseBitextMining· MTEB Multimodal Boards data docs15.3%
29nomic-embed-text-v1.5BitextMining· MTEB Multimodal Boards data docs15.2%
30Jina Embedding B EN v1BitextMining· MTEB Multimodal Boards data docs15%
31BGE Base En V1.5BitextMining· MTEB Multimodal Boards data docs14.6%
32bge-small-en-v1.5BitextMining· MTEB Multimodal Boards data docs13.5%
33All MiniLM L12 V2BitextMining· MTEB Multimodal Boards data docs12.7%
34All Mpnet Base V2BitextMining· MTEB Multimodal Boards data docs12.6%
35All MiniLM L6 V2BitextMining· MTEB Multimodal Boards data docs11.5%
36Jina Embeddings v2 Base ENBitextMining· MTEB Multimodal Boards data docs0.33%