Deep Research on BrowseComp-ZH (Pass@k)
58.1Pass@1OpenAI-o3
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OpenAI-o3Model Category=Advanced Models2025.10 | 58.1 | — | |
| DeepSeek-V3.1Model Category=Advanced Models2025.10 | 49.2 | — | |
| Claude-4-SonnetModel Category=Advanced Models2025.10 | 29.1 | — | |
| Kimi-K2Model Category=Advanced Models2025.10 | 28.8 | — | |
| Web-30B-E-GRPOModel Size=30B, Training Strategy=E-GRPO2025.10 | 26.4 | 41.2 | |
| Web-30B-GRPOModel Size=30B, Training Strategy=GRPO2025.10 | 25.7 | 38.8 | |
| WebSailorModel Size=32B2025.10 | 25.5 | — | |
| Web-30B-SFTModel Size=30B, Training Strategy=SFT2025.10 | 23.8 | 38.1 | |
| Web-7B-E-GRPOModel Size=7B, Training Strategy=E-GRPO2025.10 | 18.1 | 32.1 | |
| WebDancer-QwQ2025.10 | 18 | — | |
| Web-7B-GRPOModel Size=7B, Training Strategy=GRPO2025.10 | 17.5 | 31.5 | |
| WebSailorModel Size=7B2025.10 | 14.2 | — | |
| Web-7B-SFTModel Size=7B, Training Strategy=SFT2025.10 | 13.2 | 25.6 | |
| WebThinker-RL2025.10 | 7.3 | — | |
| R1-SearcherModel Size=7B2025.10 | 0.6 | — |