Spatial Foundation Models

Large language and vision models trained on spatial relationships, physical dynamics, and embodied interaction data. These systems understand 3D geometry, can reason about object placement and manipulation, and generate contextually appropriate spatial content and instructions for XR applications.