Web Research on BrowseComp
71.2AccuracyAgentFugue
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AgentFugueArchitecture=Multi-Agent, Backbone=DeepSeek-v4-Flash, Team size=22026.05 | 71.2 | 65 | |
| Kimi-K2.5Architecture=Single-Agent ReAct, Backbone=-2026.05 | 60.6 | — | |
| Naive-Multi-AgentArchitecture=Multi-Agent, Backbone=DeepSeek-v4-Flash, Team size=22026.05 | 59 | 57.3 | |
| DeepSeek-v4-FlashArchitecture=Single-Agent ReAct, Backbone=-2026.05 | 56.2 | 56 | |
| Swarm-Multi-AgentArchitecture=Multi-Agent, Backbone=DeepSeek-v4-Flash, Team size=22026.05 | 56.2 | 57.6 | |
| OpenAI DeepResearchArchitecture=Single-Agent DeepResearch, Backbone=-2026.05 | 51.5 | — | |
| AgentFugueArchitecture=Multi-Agent, Backbone=Qwen3.5-35B-A3B, Team size=22026.05 | 50 | 54.4 | |
| Swarm-Multi-AgentArchitecture=Multi-Agent, Backbone=Qwen3.5-35B-A3B, Team size=22026.05 | 45.2 | 49 | |
| Naive-Multi-AgentArchitecture=Multi-Agent, Backbone=Qwen3.5-35B-A3B, Team size=22026.05 | 44.5 | 48.5 | |
| GLM-4.7Architecture=Single-Agent ReAct, Backbone=-2026.05 | 43.5 | 48.7 | |
| Tongyi-DeepResearchArchitecture=Single-Agent DeepResearch, Backbone=Qwen3-30B-A3B2026.05 | 43.4 | — | |
| IterResearchArchitecture=Single-Agent DeepResearch, Backbone=Qwen3-30B-A3B2026.05 | 37.3 | — | |
| Claude-Opus-4.5Architecture=Single-Agent ReAct, Backbone=-2026.05 | 37 | — | |
| AgentFoldArchitecture=Single-Agent DeepResearch, Backbone=Qwen3-30B-A3B2026.05 | 36.2 | — | |
| Qwen3.5-35B-A3BArchitecture=Single-Agent ReAct, Backbone=-2026.05 | 36 | 45.2 | |
| WebSailorArchitecture=Single-Agent DeepResearch, Backbone=QwQ-32B2026.05 | 10.5 | — | |
| WebThinkerArchitecture=Single-Agent DeepResearch, Backbone=QwQ-32B2026.05 | 2.8 | — |