Drone-Bench - Follow: leaderboard
Metric: Follow score (%). Source: andonlabs.com. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6 | 88.6 |
| 2 | Claude Fable 5.1 | 84.3 |
| 3 | Claude Opus 5 | 80.7 |
| 4 | Claude Fable 5 | 78 |
| 5 | Gemini 3.1 Pro (Preview) | 66.2 |
| 6 | GPT-5.6 Sol | 65.2 |
| 7 | Claude Opus 4.8 | 53.1 |
| 8 | GPT-5.5 | 42.7 |
| 9 | Claude Opus 4.7 | 38.8 |
| 10 | O3 | 36.1 |
| 11 | Gemini 2.5 Pro | 23.7 |
| 12 | Claude Opus 4.5 | 20 |
| 13 | GPT-5.2 | 13.5 |
| 14 | Claude Opus 4 | 9.6 |
| 15 | O1 | 8.1 |
Interactive version: theaggregate.ai/benchmark?slug=drone-bench-follow · How It Works · Data refreshed daily, snapshot 2026-09-19.