Deep Search on BrowseComp-ZH (test)
58.1AccuracyOpenAI-o3
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI-o3Context Budget=128k2026.01 | 58.1 | |
| GLM-4.6Context Budget=128k2026.01 | 49.5 | |
| DeepSeek-v3.1Context Budget=128k2026.01 | 49.2 | |
| Tongyi-DeepResearchContext Budget=128k2026.01 | 46.7 | |
| GLM-4.5Context Budget=128k2026.01 | 37.5 | |
| DeepDive-30B-SFT + C-GRPOContext Budget=128k2026.01 | 33.3 | |
| WebExplorer-8BContext Budget=128k2026.01 | 32 | |
| DeepDive-32BContext Budget=128k2026.01 | 29.7 | |
| DeepDive-30B-SFT + GRPOContext Budget=128k2026.01 | 26.1 | |
| DeepDive-30B-SFT + C-GRPOContext Budget=64k2026.01 | 26 | |
| WebSailor-32BContext Budget=32k2026.01 | 25.5 | |
| DeepDive-4B-SFT + C-GRPOContext Budget=128k2026.01 | 24.7 | |
| DeepDive-30B-SFTContext Budget=128k2026.01 | 24.7 | |
| DeepDive-30B-SFT + GRPOContext Budget=64k2026.01 | 24.1 | |
| DeepDive-30B-SFT + E-GRPOContext Budget=128k2026.01 | 24 | |
| DeepDive-4B-SFT + E-GRPOContext Budget=128k2026.01 | 20.2 | |
| DeepDive-4B-SFT + C-GRPOContext Budget=64k2026.01 | 18.2 | |
| DeepDive-4B-SFT + GRPOContext Budget=128k2026.01 | 17.5 | |
| DeepDive-30B-SFT + E-GRPOContext Budget=64k2026.01 | 17.1 | |
| DeepDive-4B-SFTContext Budget=128k2026.01 | 16.6 | |
| DeepDive-4B-SFT + GRPOContext Budget=64k2026.01 | 16.6 | |
| DeepDive-4B-SFT + E-GRPOContext Budget=64k2026.01 | 16.5 | |
| DeepDive-30B-SFTContext Budget=64k2026.01 | 15.8 | |
| Asearcher-Web-32BContext Budget=128k2026.01 | 15.6 | |
| DeepDive-9BContext Budget=128k2026.01 | 15.1 | |
| WebSailor-7BContext Budget=32k2026.01 | 14.2 | |
| DeepDive-4B-SFTContext Budget=64k2026.01 | 10.1 |