Benchmark
MLEB - GDPR Holdings Retrieval
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.
- Models measured
- 18
- Scores recorded
- 18
- Sources
- 1
- IQ weight
- 0.67
Score distribution
- 87.3%
- Mean
- 87%
- Median
- 2.9%
- Std dev
- 83.8%
- 10th percentile
- 91.2%
- 90th percentile
- 82.9%
- Min
This benchmark has a weight of 0.67 for embedding models, measured over 18 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Gemini Embedding 001GDPR Holdings Retrieval· MLEB data | 93.8% |
| 2 | voyage-4GDPR Holdings Retrieval· MLEB data | 91.6% |
| 3 | voyage-4-largeGDPR Holdings Retrieval· MLEB data | 91.2% |
| 4 | voyage-4-liteGDPR Holdings Retrieval· MLEB data | 90.4% |
| 5 | Text Embedding 3 LargeGDPR Holdings Retrieval· MLEB data | 89.6% |
| 6 | snowflake-arctic-embed-l-v2.0GDPR Holdings Retrieval· MLEB data | 87.8% |
| 7 | snowflake-arctic-embed-m-v2.0GDPR Holdings Retrieval· MLEB data | 87.3% |
| 8 | Jina Embeddings v5 Text NanoGDPR Holdings Retrieval· MLEB data | 87.2% |
| 9 | Jina Embeddings v5 Text SmallGDPR Holdings Retrieval· MLEB data | 87% |
| 10 | Text Embedding Ada 002GDPR Holdings Retrieval· MLEB data | 86.8% |
| 11 | Qwen3 Embedding 4BGDPR Holdings Retrieval· MLEB data | 86.5% |
| 12 | Jina Embeddings v4GDPR Holdings Retrieval· MLEB data | 86.1% |
| 13 | Qwen3 Embedding 8BGDPR Holdings Retrieval· MLEB data | 85.4% |
| 14 | Text Embedding 3 SmallGDPR Holdings Retrieval· MLEB data | 85.4% |
| 15 | Multilingual E5 Large InstructGDPR Holdings Retrieval· MLEB data | 84.4% |
| 16 | Qwen3 Embedding 0.6BGDPR Holdings Retrieval· MLEB data | 83.8% |
| 17 | BGE M3GDPR Holdings Retrieval· MLEB data | 83.5% |
| 18 | Granite Embedding Small English R2GDPR Holdings Retrieval· MLEB data | 82.9% |
Data credits: MLEB (Isaacus), MIT (licence)