Reasoning on DeepSearchQA 2026 (test)
24.1ScoreDOLORES
Evaluation Results
| Method | Links | |
|---|---|---|
| DOLORESBase Model=Qwen3 32B Thinking2026.05 | 24.1 | |
| Deep ResearchBase Model=Qwen3 32B Thinking2026.05 | 23 | |
| CodeActBase Model=Qwen3 32B Thinking2026.05 | 21.3 | |
| RLMBase Model=Qwen3 32B Thinking2026.05 | 19 | |
| DOLORESBase Model=Llama-3.3 70B Instruct2026.05 | 18.7 | |
| CodeActBase Model=Llama-3.3 70B Instruct2026.05 | 18.4 | |
| ReActBase Model=Qwen3 32B Thinking2026.05 | 17.7 | |
| RLMBase Model=Llama-3.3 70B Instruct2026.05 | 17.5 | |
| DOLORESBase Model=Qwen3 8B Thinking2026.05 | 16.1 | |
| ReActBase Model=Llama-3.3 70B Instruct2026.05 | 15.5 | |
| CodeActBase Model=Qwen3 8B Thinking2026.05 | 14.9 | |
| Deep ResearchBase Model=Qwen3 8B Thinking2026.05 | 14.2 | |
| RLMBase Model=Qwen3 8B Thinking2026.05 | 13.9 | |
| Deep ResearchBase Model=Llama-3.3 70B Instruct2026.05 | 12.7 | |
| ReActBase Model=Qwen3 8B Thinking2026.05 | 12.4 |