Deep Research on BrowseComp (Pass@1, Pass@3)
50.9Pass@1OpenAI-o3
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OpenAI-o3Model Category=Advanced Models2025.10 | 50.9 | — | |
| DeepSeek-V3.1Model Category=Advanced Models2025.10 | 30 | — | |
| WebSailor-30BRL=MEM1-GRPO, Paradigm=MEM12025.09 | 19.5 | 29.7 | |
| WebSailor-30BRL=ReSum-GRPO, Paradigm=ReSum2025.09 | 18.3 | 26.5 | |
| WebExplorer-8BRL=GRPO, Paradigm=ReAct2025.09 | 15.7 | — | |
| DeepDive-32BRL=GRPO, Paradigm=ReAct2025.09 | 15.3 | — | |
| WebSailor-30BRL=GRPO, Paradigm=ReSum2025.09 | 15 | 25 | |
| WebSailor-30BRL=GRPO, Paradigm=ReAct2025.09 | 14.3 | 21.5 | |
| Kimi-K2Model Category=Advanced Models2025.10 | 14.1 | — | |
| MiroThinker-32Bv0.1RL=DPO, Paradigm=ReAct2025.09 | 13 | — | |
| Web-30B-E-GRPOModel Size=30B, Training Strategy=E-GRPO2025.10 | 12.9 | 21 | |
| WebSailor-30BRL=-, Paradigm=ReAct2025.09 | 12.8 | 21.8 | |
| Web-30B-GRPOModel Size=30B, Training Strategy=GRPO2025.10 | 12.3 | 18.9 | |
| WebSailor-7BRL=ReSum-GRPO, Paradigm=ReSum2025.09 | 12.3 | 18.5 | |
| Claude-4-SonnetModel Category=Advanced Models2025.10 | 12.2 | — | |
| Web-30B-SFTModel Size=30B, Training Strategy=SFT2025.10 | 10.8 | 18.5 | |
| WebSailorModel Size=32B2025.10 | 10.5 | — | |
| Web-7B-E-GRPOModel Size=7B, Training Strategy=E-GRPO2025.10 | 9.3 | 16.1 | |
| WebSailor-3BRL=ReSum-GRPO, Paradigm=ReSum2025.09 | 9.2 | 13 | |
| MiroThinker-8Bv0.1RL=DPO, Paradigm=ReAct2025.09 | 8.7 | — | |
| WebSailor-3BRL=GRPO, Paradigm=ReSum2025.09 | 8.5 | 13 | |
| WebSailor-7BRL=GRPO, Paradigm=ReSum2025.09 | 8.5 | 15 | |
| WebSailorModel Size=7B2025.10 | 6.7 | — | |
| Web-7B-GRPOModel Size=7B, Training Strategy=GRPO2025.10 | 6.3 | 11.7 | |
| WebSailor-7BRL=GRPO, Paradigm=ReAct2025.09 | 5.8 | 10 | |
| Web-7B-SFTModel Size=7B, Training Strategy=SFT2025.10 | 5.7 | 10.5 | |
| WebSailor-7BRL=-, Paradigm=ReAct2025.09 | 5.7 | 10.3 | |
| ASearcher-QWQ-32BRL=GRPO, Paradigm=ReAct2025.09 | 5.2 | — | |
| WebSailor-3BRL=GRPO, Paradigm=ReAct2025.09 | 4.2 | 8.5 | |
| WebDancer-QwQ2025.10 | 3.8 | — | |
| WebSailor-3BRL=-, Paradigm=ReAct2025.09 | 3.3 | 5.6 | |
| WebThinker-RL2025.10 | 2.8 | — | |
| R1-SearcherModel Size=7B2025.10 | 0.4 | — |