Skip to content

4D-fy

Text-to-4D scene generation with mixed score distillation

Visit official site

What 4d-fy is

The model is a text-to-4D generation method built on mixed score distillation sampling. It uses supervision from several pre-trained diffusion models to create scenes with structure and motion.

Main features

  • Neural parameterization of a 4D radiance field
  • Static and dynamic multi-scale hash features
  • Image and video output through volume rendering

Pros & cons

✓ What we like

  • Scenes include appearance, 3D structure and motion
  • Gradients from 3D-T2I, T2I and T2V models are combined

! What to watch out for

  • Generation starts from a text description
  • The pipeline combines several pre-trained diffusion models

FAQ

Which models contribute gradients?

A 3D-aware text-to-image model, a text-to-image model and a text-to-video model.

How are scenes rendered?

Through volume rendering of the radiance field.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review