Skip to content

Benchmark

MTEB(Europe, v1) - Retrieval

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
0.76

Score distribution

48.7%
Mean
50.9%
Median
12.2%
Std dev
38.5%
10th percentile
63.5%
90th percentile
8.6%
Min
8.6%66.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.76 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
1Qwen3 Embedding 8BRetrieval· MTEB Leaderboard data66.5%
2INF Retriever v1Retrieval· MTEB Leaderboard data66.5%
3Qwen3 Embedding 4BRetrieval· MTEB Leaderboard data65.7%
4Jina Embeddings v5 Text SmallRetrieval· MTEB Leaderboard data63.5%
5Qwen3 Embedding 0.6BRetrieval· MTEB Leaderboard data60.2%
6Text Embedding 3 LargeRetrieval· MTEB Leaderboard data59.4%
7snowflake-arctic-embed-l-v2.0Retrieval· MTEB Leaderboard data57.6%
8Multilingual E5 Large InstructRetrieval· MTEB Leaderboard data54.8%
9snowflake-arctic-embed-m-v2.0Retrieval· MTEB Leaderboard data54.5%
10Jina Embeddings v3Retrieval· MTEB Leaderboard data54%
11NV-Embed-v2Retrieval· MTEB Leaderboard data54%
12E5 Mistral 7B InstructRetrieval· MTEB Leaderboard data53.6%
13BGE M3Retrieval· MTEB Leaderboard data52.7%
14Granite Embedding 278m MultilingualRetrieval· MTEB Leaderboard data50.9%
15E5 Base V2Retrieval· MTEB Leaderboard data44.9%
16E5 Large V2Retrieval· MTEB Leaderboard data44.8%
17UAE-Large-V1Retrieval· MTEB Leaderboard data44.5%
18BGE Large En V1.5Retrieval· MTEB Leaderboard data42.3%
19Gte LargeRetrieval· MTEB Leaderboard data42.1%
20Paraphrase Multilingual MPNet v2Retrieval· MTEB Leaderboard data41.2%
21BGE Base En V1.5Retrieval· MTEB Leaderboard data40.8%
22nomic-embed-text-v1Retrieval· MTEB Leaderboard data40.7%
23Gte BaseRetrieval· MTEB Leaderboard data38.9%
24bge-small-en-v1.5Retrieval· MTEB Leaderboard data38.5%
25All MiniLM L6 V2Retrieval· MTEB Leaderboard data37.4%
26nomic-embed-text-v1.5Retrieval· MTEB Leaderboard data37.3%
27Jina Embeddings v2 Base ENRetrieval· MTEB Leaderboard data8.6%