Deep Research on Deep Research tasks (test)
3.7Interest LevelAgentic Reasoning
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Agentic Reasoningunderlying reasoning model=DeepSeek-R12025.02 | 3.7 | 4.6 | 4.2 | 4.1 | |
| Gemini-DR+2025.02 | 3.2 | 2.5 | 2.3 | 3 | |
| STORM2025.02 | 2.9 | 3.2 | 2.9 | 3.7 | |
| Search-O12025.02 | 2.5 | 2.8 | 2.1 | 3.2 | |
| RAgent2025.02 | 1.6 | 2.3 | 1.6 | 2.6 | |
| RAG2025.02 | 1.4 | 2.1 | 1.9 | 2.3 | |
| Direct Gen2025.02 | 1.2 | 1.6 | 1.2 | 1.7 |