Benchmark
Agentset Embedding Win Rate - SciFact
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages. Higher is better.
- Models measured
- 15
- Scores recorded
- 15
- Sources
- 1
- IQ weight
- 0.34
Score distribution
- 48.5%
- Mean
- 44.4%
- Median
- 12.1%
- Std dev
- 32.5%
- 10th percentile
- 62.2%
- 90th percentile
- 29.4%
- Min
This benchmark has a weight of 0.34 for embedding models, measured over 15 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 2SciFact· Agentset | 70.6% |
| 2 | voyage-4SciFact· Agentset | 62.2% |
| 3 | Text Embedding 3 LargeSciFact· Agentset | 61.3% |
| 4 | Jina Embeddings v5 Text SmallSciFact· Agentset | 59.4% |
| 5 | Embed Multilingual V3.0SciFact· Agentset | 57.5% |
| 6 | Qwen3 Embedding 0.6BSciFact· Agentset | 55.6% |
| 7 | voyage-3-largeSciFact· Agentset | 55% |
| 8 | Text Embedding 3 SmallSciFact· Agentset | 44.4% |
| 9 | Text Embedding 004SciFact· Agentset | 43.8% |
| 10 | Jina Embeddings v3SciFact· Agentset | 43.1% |
| 11 | Qwen3 Embedding 8BSciFact· Agentset | 41.3% |
| 12 | voyage-3.5-liteSciFact· Agentset | 36.3% |
| 13 | voyage-3.5SciFact· Agentset | 35.6% |
| 14 | Qwen3 Embedding 4BSciFact· Agentset | 32.5% |
| 15 | BGE M3SciFact· Agentset | 29.4% |