Information Retrieval and Question Answering on DeepSearch-QA
91.3Avg@3 ScoreClaude-4.6-Opus
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Claude-4.6-OpusAgent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 91.3 | — | — | — | — | |
| MiroThinker-H1Agent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 80.6 | — | — | — | — | |
| Claude-4.5-OpusAgent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 80 | — | — | — | — | |
| OpenAI-GPT-5Agent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 79 | — | — | — | — | |
| Seed-2.0-ProAgent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 77.4 | — | — | — | — | |
| Kimi-K2.5Agent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 77.1 | — | — | — | — | |
| Gemini-3.0-ProAgent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 76.9 | — | — | — | — | |
| MiroThinker-1.7Agent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 72.1 | — | — | — | — | |
| MiroThinker-1.7-miniAgent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 67.9 | — | — | — | — | |
| DeepSeek-V3.2Agent Style=ReAct, Max interaction turns=300, Judge LLM=gpt-4.1-2025-04-14, Max episode retries=5, Context management retention=52026.03 | 60.9 | — | — | — | — | |
| AgentOrchestraBackbone=Kimi-K22026.02 | — | 20 | 0.0263 | 437.06 | 9.9 | |
| Co-SightBackbone=Kimi-K22026.02 | — | 4 | 0.0025 | 895.88 | 2.8 | |
| Flash-SearcherBackbone=Kimi-K22026.02 | — | 22 | 0.0408 | 522.36 | 10 | |
| FlowSearchBackbone=Kimi-K22026.02 | — | 16 | 0.0109 | 351.76 | 5.5 | |
| JoyAgentBackbone=Kimi-K22026.02 | — | 28 | 0.0034 | 548.7 | 4 | |
| OAgentsBackbone=Kimi-K22026.02 | — | 28 | 0.0454 | 519.91 | 10.8 | |
| OWLBackbone=Kimi-K22026.02 | — | 30 | 0.0191 | 428.63 | 6.9 | |
| TodoEvolveBackbone=Kimi-K22026.02 | — | 42 | 0.0495 | 875.26 | 11.7 |