Deep Research on DeepResearchBench (DRB)
50.7Overall ScoreGPT-5 + Search
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GPT-5 + SearchCategory=Closed Deep Research2026.06 | 50.7 | 26.7 | 21.3 | 41 | 29.4 | |
| OpenAI Deep ResearchCategory=Closed Deep Research2026.06 | 46.9 | 46.8 | 45.2 | 49.2 | 47.1 | |
| Gemini 3 Pro + SearchCategory=Closed Deep Research2026.06 | 46.3 | 43.4 | 44.9 | 49.8 | 49 | |
| HOTE-8BCategory=Open Evolving Deep Research2026.06 | 45.9 | 44.9 | 45.4 | 47.8 | 45.8 | |
| Dr. Zero-8BCategory=Evolving Methods2026.06 | 43.7 | 41.5 | 42.1 | 46.5 | 44.7 | |
| GSPOCategory=RL Methods2026.06 | 43.6 | 42.5 | 40.9 | 47.3 | 43.7 | |
| DR Tulu-8B-RLCategory=Open Deep Research2026.06 | 43.4 | 41.7 | 41.8 | 48.2 | 41.3 | |
| GRPOCategory=RL Methods2026.06 | 43.1 | 40.8 | 42.1 | 46.9 | 42.6 | |
| REINFORCE++Category=RL Methods2026.06 | 43.1 | 41.2 | 42.7 | 46.1 | 42.4 | |
| SPICE-8BCategory=Evolving Methods2026.06 | 42.1 | 40.6 | 40.9 | 46.1 | 40.8 | |
| Tongyi DeepResearch-30B-A3BCategory=Open Deep Research Models2026.06 | 40.6 | 39.1 | 34.3 | 46.8 | 45.4 | |
| WebThinker-32B-DPO (report)Category=Fixed Pipeline Deep Research2026.06 | 40.6 | 39.4 | 35.4 | 46 | 43.5 | |
| DR Tulu-8B-SFTCategory=Open Deep Research2026.06 | 39 | 36.3 | 35.3 | 45.5 | 39.5 | |
| WebThinker QwQ-32B (report)Category=Fixed Pipeline Deep Research2026.06 | 37.9 | 36.2 | 32.6 | 43.2 | 42.9 | |
| WebExplorer-8BCategory=Open Deep Research Models2026.06 | 36.7 | 33.7 | 28.5 | 45.7 | 42.2 | |
| Ai2 ScholarQA-Claude Sonnet (report)Category=Fixed Pipeline Deep Research2026.06 | 36.1 | 35.1 | 32 | 40.5 | 38.9 | |
| WebThinker-32B-DPOCategory=Open Deep Research Models2026.06 | 23.3 | 19.7 | 12.3 | 36.8 | 26.3 | |
| Qwen3-235B-A22BCategory=Open Deep Research Models2026.06 | 22.5 | 19.1 | 17.3 | 30.6 | 25.1 | |
| Qwen3-8BCategory=Open Deep Research Models2026.06 | 18.2 | 14.3 | 8.7 | 29.5 | 24.4 | |
| Search-R1-7BCategory=Open Deep Research Models2026.06 | 9.5 | 5.2 | 2.1 | 18.6 | 16.8 | |
| ASearcher-Web-7BCategory=Open Deep Research Models2026.06 | 7.8 | 5.1 | 1.7 | 15.2 | 11.8 |