Drone-Bench is a benchmark created by Andon Labs that measures AI models' ability to autonomously program drone behavior. The model must write and adapt control code so the drone achieves given goals, combining coding skills with an understanding of a robot's physical behavior in space.
The benchmark belongs to the "robotics" part of Andon Labs' evaluation suite (alongside Butter-Bench) and tests how well models' language-and-coding abilities translate into controlling real or simulated flying platforms.
