Urban Airspace Navigation - Success Rate: leaderboard

Metric: Success rate (%) averaged over short, middle and long goal-oriented aerial navigation episodes in urban 3D scenes (5,037 samples), the model choosing discrete 3D actions from RGB views; higher is better. Source: arxiv.org. Saturation forecast: Around March 2027. 12 models tracked.

Top models

#ModelScore
1GPT-5.134
2GPT-4.124.3
3O4 Mini20.2
4Gemini 2.5 Pro16.3
5Gemini 2.5 Flash14.9
6GPT-4o14.7
7Gemini 2.0 Flash14.7
8GLM-4.6V12
9Qwen 2.5 VL 7B4.5

Interactive version: theaggregate.ai/benchmark?slug=urban-airspace-navigation-success-rate · How It Works · Data refreshed daily, snapshot 2026-10-07.