Benchmark
Agentset Reranker Win Rate
Rerank benchmarks measure how well models put the most relevant search results first. Scores are percentages. Higher is better.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 42.7%
- Mean
- 49.8%
- Median
- 11.3%
- Std dev
- 28.6%
- 10th percentile
- 57.7%
- 90th percentile
- 28.2%
- Min
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Rerank V4.0 Pro· Agentset | 57.7% |
| 2 | Qwen3 Reranker 8B· Agentset | 51.2% |
| 3 | Rerank V4.0 Fast· Agentset | 49.8% |
| 4 | Rerank V3.5· Agentset | 40.9% |
| 5 | BGE Reranker v2 M3· Agentset | 28.6% |
| 6 | Jina Reranker v2 Base Multilingual· Agentset | 28.2% |