Skip to content

Benchmark

MTEB(Scandinavian, 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
29
Scores recorded
29
Sources
1
IQ weight
0.67

Score distribution

55.3%
Mean
61.9%
Median
14.2%
Std dev
33%
10th percentile
70.9%
90th percentile
27.5%
Min
27.5%77.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.67 for embedding models, measured over 29 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini Embedding 001Retrieval· MTEB Leaderboard data77.2%
2Text Embedding 3 LargeRetrieval· MTEB Leaderboard data73.4%
3Qwen3 Embedding 8BRetrieval· MTEB Leaderboard data71.2%
4INF Retriever v1Retrieval· MTEB Leaderboard data70.9%
5Jina Embeddings v5 Omni NanoRetrieval· MTEB Leaderboard data68.5%
6Qwen3 Embedding 4BRetrieval· MTEB Leaderboard data68.4%
7Jina Embeddings v5 Omni SmallRetrieval· MTEB Leaderboard data68%
8Jina Embeddings v5 Text SmallRetrieval· MTEB Leaderboard data68%
9Text Embedding 3 SmallRetrieval· MTEB Leaderboard data66.5%
10snowflake-arctic-embed-l-v2.0Retrieval· MTEB Leaderboard data65.4%
11Multilingual E5 LargeRetrieval· MTEB Leaderboard data64.5%
12BGE M3Retrieval· MTEB Leaderboard data64%
13Multilingual E5 Large InstructRetrieval· MTEB Leaderboard data63.9%
14E5 Mistral 7B InstructRetrieval· MTEB Leaderboard data61.9%
15Qwen3 Embedding 0.6BRetrieval· MTEB Leaderboard data61.9%
16Granite Embedding 278m MultilingualRetrieval· MTEB Leaderboard data55.8%
17Paraphrase Multilingual MPNet v2Retrieval· MTEB Leaderboard data50.4%
18E5 Base V2Retrieval· MTEB Leaderboard data48.6%
19E5 Large V2Retrieval· MTEB Leaderboard data47.1%
20Gte LargeRetrieval· MTEB Leaderboard data46.5%
21BGE Base En V1.5Retrieval· MTEB Leaderboard data45.6%
22BGE Large En V1.5Retrieval· MTEB Leaderboard data45%
23UAE-Large-V1Retrieval· MTEB Leaderboard data43.6%
24bge-small-en-v1.5Retrieval· MTEB Leaderboard data42.6%
25Gte BaseRetrieval· MTEB Leaderboard data42.2%
26All Mpnet Base V2Retrieval· MTEB Leaderboard data33%
27All MiniLM L12 V2Retrieval· MTEB Leaderboard data32%
28Jina Embedding B EN v1Retrieval· MTEB Leaderboard data29.6%
29All MiniLM L6 V2Retrieval· MTEB Leaderboard data27.5%