Deep Information Retrieval and Research on xbench DeepSearch
77.8Avg@8OpenAI-GPT-5-high
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI-GPT-5-highType=Foundation Models with Tools2025.11 | 77.8 | |
| MiroThinker-v1.0-72BParameters=72B2025.11 | 77.8 | |
| Tongyi-DeepResearch-30BType=Research Agents2025.11 | 75 | |
| Minimax-M2Type=Foundation Models with Tools2025.11 | 72 | |
| DeepSeek-V3.1Type=Foundation Models with Tools2025.11 | 71 | |
| DeepSeek-V3.2Type=Foundation Models with Tools2025.11 | 71 | |
| MiroThinker-v1.0-30BParameters=30B2025.11 | 70.6 | |
| GLM-4.6Type=Foundation Models with Tools2025.11 | 70 | |
| Kimi-ResearcherType=Research Agents2025.11 | 69 | |
| OpenAI-o3Type=Foundation Models with Tools2025.11 | 67 | |
| Claude-4.5-SonnetType=Foundation Models with Tools2025.11 | 66 | |
| Claude-4-SonnetType=Foundation Models with Tools2025.11 | 64.6 | |
| DeepMiner-32B-RLType=Research Agents2025.11 | 62 | |
| Kimi-K2-0905Type=Foundation Models with Tools2025.11 | 61 | |
| MiroThinker-v1.0-8BParameters=8B2025.11 | 60.6 | |
| WebExplorer-8B-RLType=Research Agents2025.11 | 53.7 |