SearchAgent - Bamboogle: leaderboard
SearchAgent Bamboogle slice measures exact match on deliberately hard search questions.
Metric: Exact Match (%). Source: huggingface.co. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | ReSeek-Qwen2.5-7b-Instruct | 39.2 | |
| 2 | Search-R1-Qwen2.5-7b-Instruct | 36.8 | |
| 3 | ReSeek-Qwen2.5-3b-Instruct | 30.4 | |
| 4 | Search-o1-Qwen2.5-7b-Instruct | 29.6 | |
| 5 | R1-Qwen2.5-7b-Instruct | 29.3 | |
| 6 | ZeroSearch-Qwen2.5-7b-Instruct | 27.8 | |
| 7 | Search-R1-Qwen2.5-3b-Instruct | 26.4 | |
| 8 | CoT-Qwen2.5-7b-Instruct | 23.2 | |
| 9 | RAG-Qwen2.5-7b-Instruct | 20.8 | |
| 10 | Direct-Inference-Qwen2.5-7b-Instruct | 12 |
Interactive version: theaggregate.ai/benchmark?slug=searchagent-bamboogle · How It Works · Data refreshed daily, snapshot 2026-10-20.