Web Browsing Competition on Browse Comp
67.6ScoreDeepSeek-V3.2
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-V3.2Model Access Type=Closed-Source2026.02 | 67.6 | |
| Seed-1.8Model Access Type=Closed-Source2026.02 | 67.6 | |
| OpenAI-GPT-5-highModel Access Type=Closed-Source2026.02 | 54.9 | |
| GLM-4.6Model Access Type=Closed-Source2026.02 | 45.1 | |
| Minimax-M2Model Access Type=Closed-Source2026.02 | 44 | |
| Tongyi-DeepResearch 30BModel Access Type=Open-Source2026.02 | 43.4 | |
| WebLeaper-30B-A3BModel Access Type=Open-Source2026.02 | 38.8 | |
| IterResearch-30B-A3BModel Access Type=Open-Source2026.02 | 37.3 | |
| WebSailor-V2-30B-A3BModel Access Type=Open-Source2026.02 | 35.3 | |
| MiroThinker 8BModel Access Type=Open-Source2026.02 | 31.1 | |
| AgentCPM-Explore-4BModel Access Type=Open-Source2026.02 | 24.1 | |
| Merged-Model-4BModel Access Type=Open-Source2026.02 | 21.9 | |
| Claude-4.5-SonnetModel Access Type=Closed-Source2026.02 | 19.6 | |
| WebDancer-QwQ-32BModel Access Type=Open-Source2026.02 | 3.8 |