Open-Ended Deep Research on DeepConsult
64.42Win RateDualGraph
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DualGraphBackbone=GPT-4.1, Knowledge Graph usage=true2026.02 | 64.42 | 23.08 | 12.5 | 6.42 | |
| Gemini-2.5-Pro-DeepResearch2026.02 | 61.27 | 31.13 | 7.6 | 6.7 | |
| DualGraphBackbone=GPT-4.1, Knowledge Graph usage=false2026.02 | 57.61 | 28.26 | 14.13 | 6.24 | |
| Doubao-Research2026.02 | 29.95 | 40.35 | 29.7 | 5.42 | |
| Claude-Research2026.02 | 25 | 38.89 | 36.11 | 4.6 | |
| LangChain-DeepResearchBackbone=GPT-4.12026.02 | 14.46 | 44.6 | 40.9 | 4.26 | |
| Tongyi-DeepResearchBackbone=GPT-4.12026.02 | 13.48 | 10.54 | 75.98 | 2.63 | |
| Qwen-AgentBackbone=GPT-4.12026.02 | 4.41 | 4.41 | 91.17 | 1.99 | |
| OpenAI-DeepResearch2026.02 | 0 | 100 | 0 | 5 |