Skip to content

Benchmark

ChemTEB(v1.1) - BitextMining

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
0.64

Score distribution

22.8%
Mean
21%
Median
13.2%
Std dev
10%
10th percentile
39%
90th percentile
6%
Min
6%60%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.64 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 LargeBitextMining· MTEB Extra Text Boards data60%
2Qwen3 Embedding 8BBitextMining· MTEB Extra Text Boards data52%
3Jina Embeddings v5 Text SmallBitextMining· MTEB Extra Text Boards data41%
4Jina Embeddings v5 Omni NanoBitextMining· MTEB Extra Text Boards data39%
5Qwen3 Embedding 4BBitextMining· MTEB Extra Text Boards data34%
6Text Embedding Ada 002BitextMining· MTEB Extra Text Boards data34%
7Titan Embeddings G1 - TextBitextMining· MTEB Extra Text Boards data30%
8nomic-embed-text-v1.5BitextMining· MTEB Extra Text Boards data26%
9Embed Multilingual V3.0BitextMining· MTEB Extra Text Boards data24%
10Text Embedding 3 SmallBitextMining· MTEB Extra Text Boards data24%
11E5 Large V2BitextMining· MTEB Extra Text Boards data23%
12nomic-embed-text-v1BitextMining· MTEB Extra Text Boards data23%
13Multilingual E5 LargeBitextMining· MTEB Extra Text Boards data21%
14Titan Text Embeddings V2BitextMining· MTEB Extra Text Boards data21%
15BGE Base En V1.5BitextMining· MTEB Extra Text Boards data18%
16BGE Large En V1.5BitextMining· MTEB Extra Text Boards data17%
17Multilingual E5 Large InstructBitextMining· MTEB Extra Text Boards data16%
18All MiniLM L6 V2BitextMining· MTEB Extra Text Boards data15%
19E5 Base V2BitextMining· MTEB Extra Text Boards data14%
20All MiniLM L12 V2BitextMining· MTEB Extra Text Boards data13%
21bge-large-enBitextMining· MTEB Extra Text Boards data13%
22bge-small-en-v1.5BitextMining· MTEB Extra Text Boards data13%
23Embed English V3.0BitextMining· MTEB Extra Text Boards data12%
24Qwen3 Embedding 0.6BBitextMining· MTEB Extra Text Boards data10%
25All Mpnet Base V2BitextMining· MTEB Extra Text Boards data9%
26BGE M3BitextMining· MTEB Extra Text Boards data7%
27Multi Qa Mpnet Base Dot V1BitextMining· MTEB Extra Text Boards data6%