Web-based research on DeepResearchBench (test)
52.91AccuracyEvoAgent
Evaluation Results
| Method | Links | |
|---|---|---|
| EvoAgentMeta-agent=GPT-5.4-Nano2026.06 | 52.91 | |
| Skill-MAS-optimizedMeta-agent=DeepSeek-V4-Flash2026.06 | 51.69 | |
| Skill-MAS-optimizedMeta-agent=Qwen3.5-Plus2026.06 | 51.28 | |
| AFlowMeta-agent=DeepSeek-V4-Flash2026.06 | 49.93 | |
| EvoAgentMeta-agent=DeepSeek-V4-Flash2026.06 | 49.1 | |
| Skill-MAS-optimizedMeta-agent=GPT-5.4-Nano2026.06 | 48.9 | |
| Skill-MAS-initMeta-agent=Qwen3.5-Plus2026.06 | 48.88 | |
| EvoAgentMeta-agent=Qwen3.5-Plus2026.06 | 47.02 | |
| AFlowMeta-agent=Qwen3.5-Plus2026.06 | 46.17 | |
| Skill-MAS-initMeta-agent=DeepSeek-V4-Flash2026.06 | 44.99 | |
| Skill-MAS-optimizedMeta-agent=Gemini-3.1-Flash2026.06 | 44.71 | |
| AOrchestraMeta-agent=GPT-5.4-Nano2026.06 | 44.68 | |
| AFlowMeta-agent=GPT-5.4-Nano2026.06 | 44.29 | |
| AOrchestraMeta-agent=Qwen3.5-Plus2026.06 | 44.05 | |
| MAS-OrchestraMeta-agent=GPT-5.4-Nano2026.06 | 43.83 | |
| MAS2Meta-agent=GPT-5.4-Nano2026.06 | 43.2 | |
| AFlowMeta-agent=Gemini-3.1-Flash2026.06 | 42.31 | |
| AOrchestraMeta-agent=DeepSeek-V4-Flash2026.06 | 41.03 | |
| MAS2Meta-agent=Qwen3.5-Plus2026.06 | 40.87 | |
| EvoAgentMeta-agent=Gemini-3.1-Flash2026.06 | 40.76 | |
| MAS-OrchestraMeta-agent=DeepSeek-V4-Flash2026.06 | 40.63 | |
| Skill-MAS-initMeta-agent=GPT-5.4-Nano2026.06 | 40.45 | |
| MAS-OrchestraMeta-agent=Qwen3.5-Plus2026.06 | 40.43 | |
| MAS-OrchestraMeta-agent=Gemini-3.1-Flash2026.06 | 40.4 | |
| AOrchestraMeta-agent=Gemini-3.1-Flash2026.06 | 39.49 | |
| MAS2Meta-agent=Gemini-3.1-Flash2026.06 | 38.03 | |
| Skill-MAS-initMeta-agent=Gemini-3.1-Flash2026.06 | 36.1 | |
| MAS2Meta-agent=DeepSeek-V4-Flash2026.06 | 35.31 |