Drone-Bench - Navigate: leaderboard
Drone-Bench Navigate: write a control loop that plans a collision-free path on an obstacle map and flies to a target room while calling a delayed, intermittently failing localizer. Scored on reaching the target, path optimality and speed; average over runs of each run's best submission, as a percentage (human baseline 92.0).
Metric: Navigate score (%). Source: andonlabs.com. Saturation forecast: Around February 2027. 19 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 98.2 | #6 |
| 2 | Claude Opus 5.5 | 96.71 | #3 |
| 3 | Claude Opus 4.8 | 93.69 | #39 |
| 4 | GPT-5.6 Sol | 91.73 | #22 |
| 5 | Claude Opus 4.7 | 89.13 | #51 |
| 6 | GPT-5.5 | 88.36 | #33 |
| 7 | Claude Fable 5.1 | 85.18 | #10 |
| 8 | Claude Opus 5 | 84.73 | #16 |
| 9 | Claude Opus 4.5 | 82.19 | #87 |
| 10 | Gemini 3.1 Pro (Preview) | 78.28 | #64 |
| 11 | GPT-5.2 | 78.26 | #115 |
| 12 | Claude Fable 5 | 68.54 | #20 |
| 13 | GPT-5 | 61.87 | #105 |
| 14 | O3 | 56.19 | #132 |
| 15 | Gemini 2.5 Pro | 48.57 | #159 |
No result here: #7 GPT-6.1 Sol, #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=drone-bench-navigate · How It Works · Data refreshed daily, snapshot 2026-11-05.