All it needs to start is a single image prompt either generated by AI or from a real-world photo. There are limitations: DeepMind says the model can generate “consistent” worlds for up to 60 ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback