Skip to content

PhysGen

Physics-based image-to-video generation

Visit official site

What PhysGen is

The method is described as an image-to-video approach that turns one image plus an input condition such as force or torque into realistic, physically plausible video. It combines model-based physics simulation with data-driven video generation.

What you can do with it

  • Turn one image into a video with a condition
  • Apply a force or torque to objects
  • Control the generated motion precisely

Getting started

  • Visit the PhysGen site
  • Upload a static image
  • Choose a condition such as force or torque

Pros & cons

✓ What we like

  • Generated motion is described as precisely controllable
  • Physics and appearance are both described as realistic

! What to watch out for

  • An input condition such as force or torque is required
  • The pipeline starts from a single static image

FAQ

What input is needed?

A single image plus a condition such as force or torque applied to objects.

What does it produce?

The material describes realistic, physically plausible and time-coherent video.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review