vorstellar labs

is a research lab building generative world models to derive visual cognitive intelligence from large-scale video and synthetic data, enabling machines to conceptualize physical environments like humans.

our team includes cognitive scientists from Stanford, robotics researchers from UCSD, and NVIDIA inference engineers, collaborating closely to bridge theory and scalable implementation.

we are backed by Y Combinator and select angel investors, including the CEOs of Average Database and SD Compute.