ResearcherBench — leaderboard
Evaluates deep AI research systems on producing comprehensive research reports, measuring coverage of key insights, faithfulness to sources, and groundedness of claims.
Metric: Coverage (%). Source: github.com. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | OpenAI Deep Research | 70.3 |
| 2 | Gemini Deep Research | 69.3 |
| 3 | Claude Research | 61.1 |
| 4 | Mita Deep Research | 58.4 |
| 5 | Doubao Deep Research | 57.5 |
| 6 | Perplexity Deep Research | 48 |
| 7 | Perplexity: Sonar Reasoning Pro | 46.6 |
| 8 | Grok3 DeepSearch | 44.1 |
| 9 | Grok3 DeeperSearch | 44 |
| 10 | GPT-4o Search Preview | 35.8 |
Interactive version: theaggregate.ai/benchmark?slug=researcherbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.