FutureSearch DRB - Gather Evidence: leaderboard
FutureSearch Deep Research Bench task slice for gathering supporting evidence from web sources.
Metric: Average Score. Source: evals.futuresearch.ai. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4 | 0.4 | #163 |
| 2 | GPT-5 | 0.38 | #98 |
| 3 | O3 | 0.38 | #124 |
| 4 | Gemini 3 Pro | 0.36 | #80 |
| 5 | Claude Opus 4.1 | 0.36 | #140 |
| 6 | Claude Sonnet 4 | 0.34 | #199 |
| 7 | Grok 4 | 0.34 | #181 |
| 8 | Claude Sonnet 4.6 | 0.33 | #90 |
| 9 | GPT-5.1 | 0.33 | #135 |
| 10 | Claude Opus 4.5 | 0.32 | #83 |
| 11 | Claude Opus 4.6 | 0.32 | #63 |
| 12 | Gemini 3 Flash | 0.3 | #99 |
| 13 | Claude Haiku 4.5 | 0.29 | #280 |
| 14 | Gemini 2.5 Pro | 0.28 | #152 |
| 15 | Claude Sonnet 4.5 | 0.28 | #142 |
Interactive version: theaggregate.ai/benchmark?slug=futuresearch-drb-gather-evidence · How It Works · Data refreshed daily, snapshot 2026-10-20.