Deep Research on Xbench DeepResearch
67AccuracyOpenAI o4-mini
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI o4-minitype=Foundation Models with Tools2026.03 | 67 | |
| OPENRESEARCHER2026.03 | 65 | |
| Claude-4-Sonnettype=Foundation Models with Tools2026.03 | 64 | |
| DeepSeek-R1type=Foundation Models with Tools2026.03 | 55 | |
| Nemotron-3-Nanotype=Foundation Models with Tools2026.03 | 55 | |
| WebSailor-72Btype=Deep Research Agents2026.03 | 55 | |
| DeepMiner-32Btype=Deep Research Agents2026.03 | 53 | |
| Kimi-K2type=Foundation Models with Tools2026.03 | 50 | |
| Live-EvoBase Model=GPT-4.1-mini2026.02 | 46 | |
| MiroFlowBase Model=GPT-4.1-mini2026.02 | 45 | |
| Qwen-DeepResearchBase Model=GPT-4.1-mini2026.02 | 43 | |
| ASearcher-QwQ-32Btype=Deep Research Agents2026.03 | 42 | |
| ReMemBase Model=GPT-4.1-mini2026.02 | 40 | |
| WebDancer-QwQ-32Btype=Deep Research Agents2026.03 | 39 |