Web Browsing and Interaction on Browsecomp
51.5AccuracyOpenAI DeepResearch
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| OpenAI DeepResearchCategory=Close-sourced System2026.02 | 51.5 | — | — | — | |
| OpenAI o3Category=Close-sourced System2026.02 | 49.7 | — | — | — | |
| WebClipper (Hybrid)Category=Pruning Method, Base Model=Tongyi-DeepResearch, Variant=Hybrid2026.02 | 46.7 | 0.428 | 60.42 | 11,507 | |
| WebClipper (Eff)Category=Pruning Method, Base Model=Tongyi-DeepResearch, Variant=Efficiency2026.02 | 42.7 | 0.431 | 56.5 | 10,599 | |
| Tongyi-DeepResearchCategory=Open-sourced Agent, Evaluation Environment=Unified2026.02 | 41 | 0.385 | 63.7 | 12,014 | |
| Prompt ControlCategory=Pruning Method, Base Model=Tongyi-DeepResearch2026.02 | 37.3 | 0.372 | 62.8 | 12,222 | |
| Coarse PruneCategory=Pruning Method, Base Model=Tongyi-DeepResearch2026.02 | 22 | 0.326 | 37.1 | 8,365 | |
| DeepSeek-R1-671BCategory=Open-sourced Agent, Evaluation Environment=Unified2026.02 | 14.4 | 0.248 | 10.25 | 2,022 | |
| WebExplorerCategory=Open-sourced Agent2026.02 | 13.7 | 0.229 | 29.43 | 6,289 | |
| Claude-4-SonnetCategory=Close-sourced System2026.02 | 12.2 | — | — | — | |
| Kimi-K2-Instruct-0905Category=Open-sourced Agent, Evaluation Environment=Unified2026.02 | 9.4 | 0.169 | 16.65 | 3,426 | |
| Qwen3-235B-A22B-Instruct-2507Category=Open-sourced Agent, Evaluation Environment=Unified2026.02 | 4.6 | 0.087 | 13.7 | 1,837 |