Drone-Bench - Detect: leaderboard
Metric: Detect score (%). Source: andonlabs.com. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 | 87.8 |
| 2 | Claude Fable 5.1 | 85.7 |
| 3 | Claude Opus 4.7 | 84.2 |
| 4 | Gemini 3.1 Pro (Preview) | 82.4 |
| 5 | Claude Opus 4.8 | 81 |
| 6 | Claude Opus 5 | 78.5 |
| 7 | Claude Opus 4.5 | 69.8 |
| 8 | GPT-6 | 67.1 |
| 9 | GPT-5.5 | 63.5 |
| 10 | GPT-5.6 Sol | 61.3 |
| 11 | O1 | 58.7 |
| 12 | GPT-5.2 | 56.1 |
| 13 | O3 | 55.4 |
| 14 | GPT-4o (2024-11-20) | 55.3 |
| 15 | Claude Opus 4 | 52.9 |
Interactive version: theaggregate.ai/benchmark?slug=drone-bench-detect · How It Works · Data refreshed daily, snapshot 2026-09-19.