Deep Research on POP (test)
87.9Mean Correct RateWebSailor
Evaluation results
| Method | Links | |
|---|---|---|
| WebSailorType=Live web2025.10 | 87.9 | |
| Our best agentType=Live web2025.10 | 86.3 | |
| Our base agentType=Live web2025.10 | 83.6 | |
| ASearcherType=Live web2025.10 | 81.9 | |
| DeepResearcherType=Live web2025.10 | 81.05 | |
| Search-R1Type=Static & simulated web2025.10 | 77.2 | |
| R1-SearcherType=Static & simulated web2025.10 | 65.1 | |
| ZeroSearchType=Static & simulated web2025.10 | 39.7 |