Benchmark
Agentset Embedding Elo - SciFact
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Higher scores are better.
- Models measured
- 15
- Scores recorded
- 15
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 1,500
- Mean
- 1,500
- Median
- 0
- Std dev
- 1,500
- 10th percentile
- 1,500
- 90th percentile
- 1,500
- Min
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | BGE M3SciFact· Agentset | 1,500 |
| 2 | Embed Multilingual V3.0SciFact· Agentset | 1,500 |
| 3 | Gemini Embedding 2SciFact· Agentset | 1,500 |
| 4 | Jina Embeddings v3SciFact· Agentset | 1,500 |
| 5 | Jina Embeddings v5 Text SmallSciFact· Agentset | 1,500 |
| 6 | Qwen3 Embedding 0.6BSciFact· Agentset | 1,500 |
| 7 | Qwen3 Embedding 4BSciFact· Agentset | 1,500 |
| 8 | Qwen3 Embedding 8BSciFact· Agentset | 1,500 |
| 9 | Text Embedding 004SciFact· Agentset | 1,500 |
| 10 | Text Embedding 3 LargeSciFact· Agentset | 1,500 |
| 11 | Text Embedding 3 SmallSciFact· Agentset | 1,500 |
| 12 | voyage-3-largeSciFact· Agentset | 1,500 |
| 13 | voyage-3.5SciFact· Agentset | 1,500 |
| 14 | voyage-3.5-liteSciFact· Agentset | 1,500 |
| 15 | voyage-4SciFact· Agentset | 1,500 |