Web Research on BrowseComp-EN 200
14.5Pass@1THINKMERGE
Evaluation Results
| Method | Links | |
|---|---|---|
| THINKMERGEmodel=WebSailor-32B, N=82025.12 | 14.5 | |
| THINKMERGEmodel=WebSailor-7B, N=42025.12 | 13.6 | |
| THINKMERGEmodel=WebSailor-32B, N=42025.12 | 13.4 | |
| THINKMERGEmodel=WebSailor-7B, N=82025.12 | 13.1 | |
| THINKMERGEmodel=WebSailor-32B, N=22025.12 | 13.1 | |
| Claude-4-Sonnet2025.12 | 12.2 | |
| WebSailor-32Bvariant=Baseline2025.12 | 11.8 | |
| THINKMERGEmodel=WebSailor-7B, N=22025.12 | 11 | |
| MiroThinker-8B-DPO-v0.1parameters=8B2025.12 | 8.7 | |
| FoldAct-7Bp_drop=0, parameters=7B2025.12 | 8.4 | |
| FoldAct-7Bp_drop=0.5, parameters=7B2025.12 | 8.3 | |
| WebSailor-7Bparameters=7B2025.12 | 6.7 | |
| THINKMERGEmodel=WebSailor-3B, N=22025.12 | 6.3 | |
| WebSailor-7Bvariant=Baseline2025.12 | 6.3 | |
| ASearcher-Web-QwQparameters=32B2025.12 | 5.2 | |
| WebSailor-3Bvariant=Baseline2025.12 | 4.7 | |
| THINKMERGEmodel=WebSailor-3B, N=42025.12 | 3.5 | |
| THINKMERGEmodel=WebSailor-3B, N=82025.12 | 2.5 | |
| GPT-4.1-mini2025.12 | 2.1 |