Web Search on xbench
66Average ScoreOpenAI-o3
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI-o32026.01 | 66 | |
| Claude-4-Sonnet2026.01 | 64 | |
| Reagent-UBackbone=Qwen3-8B2026.01 | 43 | |
| Reagent-RBackbone=Qwen3-8B2026.01 | 41 | |
| Search-o1Backbone=QwQ-32B-Preview2026.01 | 40 | |
| WebDancerBackbone=Qwen2.5-32B2026.01 | 38 | |
| DeepSeek-R1-671BBackbone=DeepSeek-R1-671B2026.01 | 32 | |
| ARPOBackbone=Qwen3-14B2026.01 | 32 | |
| Reagent w/o Agent-RRMBackbone=Qwen3-8B2026.01 | 32 | |
| ARPOBackbone=Qwen3-8B2026.01 | 25 | |
| Atom-SearcherBackbone=Qwen2.5-7B2026.01 | 21 | |
| Reagent-CBackbone=Qwen3-8B, Inference Type=Direct Inference2026.01 | 15 | |
| WebThinkerBackbone=Qwen3-8B2026.01 | 13 | |
| QwQ-32BBackbone=QwQ-32B2026.01 | 10 | |
| Qwen3-8BBackbone=Qwen3-8B2026.01 | 9 |