← Catalog
World ModelSpAItial · 2025
SpAItial
Spatial Foundation Models that generate interactive, photorealistic 3D environments from a text prompt or image.
Summary
SpAItial (Munich; founded by Matthias Niessner with Ricardo Martin-Brualla and David Novotny) is building Spatial Foundation Models — a class of generative models that reason about space and time to produce fully interactive, photorealistic 3D environments from text or images, renderable from novel viewpoints. Raised a $13M seed; early/closed.
Metadata
- Organization
- SpAItial (Munich / TUM)
- Year
- 2025
- Focus
- Spatial Foundation Models (text → interactive 3D)
- Availability
- Closed (seed stage)
Relationships
Competes with
#spatial-intelligence#3d#generative