Figure’s Helix Model Powers Home Robot Tasks

Figure’s CEO Brett Adcock announced Helix, a Vision-Language-Action model for humanoid robots, on February 20, 2025.

Helix processes visual data and natural language to control Figure 02 robots in real time, enabling tasks like handing over or placing items.

It understands new household objects without extensive programming.

Home testing begins in 2025, tackling complex, unstructured home environments.