Benchmark
MLEB - License TL;DR 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
- 0.87
Score distribution
- 67.2%
- Mean
- 66.7%
- Median
- 6.8%
- Std dev
- 59.9%
- 10th percentile
- 76.4%
- 90th percentile
- 54.1%
- Min
This benchmark has a weight of 0.87 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 | voyage-4-largeLicense TL, DR Retrieval· MLEB data | 81.4% |
| 2 | voyage-4License TL, DR Retrieval· MLEB data | 77.6% |
| 3 | Qwen3 Embedding 4BLicense TL, DR Retrieval· MLEB data | 76.4% |
| 4 | Qwen3 Embedding 8BLicense TL, DR Retrieval· MLEB data | 73.3% |
| 5 | voyage-4-liteLicense TL, DR Retrieval· MLEB data | 71.8% |
| 6 | Jina Embeddings v5 Text SmallLicense TL, DR Retrieval· MLEB data | 71% |
| 7 | Gemini Embedding 001License TL, DR Retrieval· MLEB data | 69.1% |
| 8 | Jina Embeddings v5 Text NanoLicense TL, DR Retrieval· MLEB data | 67.6% |
| 9 | Text Embedding 3 LargeLicense TL, DR Retrieval· MLEB data | 66.7% |
| 10 | Qwen3 Embedding 0.6BLicense TL, DR Retrieval· MLEB data | 64.4% |
| 11 | snowflake-arctic-embed-m-v2.0License TL, DR Retrieval· MLEB data | 64.3% |
| 12 | Jina Embeddings v4License TL, DR Retrieval· MLEB data | 64.1% |
| 13 | Granite Embedding Small English R2License TL, DR Retrieval· MLEB data | 63.6% |
| 14 | BGE M3License TL, DR Retrieval· MLEB data | 63% |
| 15 | snowflake-arctic-embed-l-v2.0License TL, DR Retrieval· MLEB data | 62.4% |
| 16 | Text Embedding 3 SmallLicense TL, DR Retrieval· MLEB data | 59.9% |
| 17 | Text Embedding Ada 002License TL, DR Retrieval· MLEB data | 59.2% |
| 18 | Multilingual E5 Large InstructLicense TL, DR Retrieval· MLEB data | 54.1% |
Data credits: MLEB (Isaacus), MIT (licence)