Report Synthesis on DeepResearch Bench (DRB)
50.6ScoreClaude Opus 4.5
Evaluation results
| Method | Links | |
|---|---|---|
| Claude Opus 4.52026.05 | 50.6 | |
| Gemini 3 Pro2026.05 | 49.6 | |
| Quest-35B2026.05 | 48.2 | |
| OpenAI-DR2026.05 | 47 | |
| Quest-30B2026.05 | 45.3 | |
| DR Tulu2026.05 | 43.4 | |
| Tongyi-DR2026.05 | 40.5 | |
| OpenResearcherEvaluation Implementation=Authors' implementation2026.05 | 35.4 |