Benchmark
MLEB - SCALR
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
- 0.80
Score distribution
- 68.1%
- Mean
- 67.8%
- Median
- 6.5%
- Std dev
- 61.5%
- 10th percentile
- 75.6%
- 90th percentile
- 55.2%
- Min
This benchmark has a weight of 0.80 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-largeSCALR· MLEB data | 79.2% |
| 2 | voyage-4SCALR· MLEB data | 77.1% |
| 3 | Qwen3 Embedding 8BSCALR· MLEB data | 75.6% |
| 4 | voyage-4-liteSCALR· MLEB data | 74.5% |
| 5 | Text Embedding 3 LargeSCALR· MLEB data | 74.5% |
| 6 | Gemini Embedding 001SCALR· MLEB data | 72.1% |
| 7 | Qwen3 Embedding 4BSCALR· MLEB data | 71.1% |
| 8 | Jina Embeddings v4SCALR· MLEB data | 69.5% |
| 9 | Jina Embeddings v5 Text SmallSCALR· MLEB data | 67.8% |
| 10 | Jina Embeddings v5 Text NanoSCALR· MLEB data | 66.6% |
| 11 | Qwen3 Embedding 0.6BSCALR· MLEB data | 66.3% |
| 12 | snowflake-arctic-embed-m-v2.0SCALR· MLEB data | 65.5% |
| 13 | Text Embedding 3 SmallSCALR· MLEB data | 65.3% |
| 14 | snowflake-arctic-embed-l-v2.0SCALR· MLEB data | 63.3% |
| 15 | Granite Embedding Small English R2SCALR· MLEB data | 62.1% |
| 16 | Multilingual E5 Large InstructSCALR· MLEB data | 61.5% |
| 17 | Text Embedding Ada 002SCALR· MLEB data | 58.2% |
| 18 | BGE M3SCALR· MLEB data | 55.2% |
Data credits: MLEB (Isaacus), MIT (licence)