Benchmark
BuiltBench(eng) - Retrieval
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.
- Models measured
- 16
- Scores recorded
- 16
- Sources
- 1
- IQ weight
- 0.71
Score distribution
- 63.6%
- Mean
- 64.6%
- Median
- 3.4%
- Std dev
- 58.8%
- 10th percentile
- 67.3%
- 90th percentile
- 57.3%
- Min
This benchmark has a weight of 0.71 for embedding models, measured over 16 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 | Qwen3 Embedding 8BRetrieval· MTEB Extra Text Boards data | 71% |
| 2 | Qwen3 Embedding 4BRetrieval· MTEB Extra Text Boards data | 67.3% |
| 3 | UAE-Large-V1Retrieval· MTEB Extra Text Boards data | 66.4% |
| 4 | Gte LargeRetrieval· MTEB Extra Text Boards data | 66.1% |
| 5 | snowflake-arctic-embed-m-v2.0Retrieval· MTEB Extra Text Boards data | 65% |
| 6 | BGE Large En V1.5Retrieval· MTEB Extra Text Boards data | 64.9% |
| 7 | E5 Large V2Retrieval· MTEB Extra Text Boards data | 64.6% |
| 8 | Multilingual E5 Large InstructRetrieval· MTEB Extra Text Boards data | 64.6% |
| 9 | snowflake-arctic-embed-l-v2.0Retrieval· MTEB Extra Text Boards data | 64.4% |
| 10 | Multilingual E5 LargeRetrieval· MTEB Extra Text Boards data | 63.1% |
| 11 | Jina Embeddings v3Retrieval· MTEB Extra Text Boards data | 62.6% |
| 12 | BGE Base En V1.5Retrieval· MTEB Extra Text Boards data | 61.6% |
| 13 | E5 Base V2Retrieval· MTEB Extra Text Boards data | 61% |
| 14 | All MiniLM L12 V2Retrieval· MTEB Extra Text Boards data | 58.8% |
| 15 | bge-small-en-v1.5Retrieval· MTEB Extra Text Boards data | 58.8% |
| 16 | BGE M3Retrieval· MTEB Extra Text Boards data | 57.3% |