Skip to content

DragNUWA

Drag an image to steer camera and object motion in video

Visit official site

What DragNUWA is

The tool is described as a video generation tool that turns direct actions on a background or image into camera motion or object motion. It generates the matching video from the dragged actions.

What you can do with it

  • Generate a finely controlled video through drag actions
  • Animate an image along a set path with stable video diffusion
  • Control video generation with text, an image and a trajectory
  • Clone the repository and download pretrained models for desktop drag generation

Reference cases

  • A user makes an art video that changes along a set path with the 1.5 release
  • A user generates a creative video with specific scene and time features through text, image and trajectory control
  • A user turns a picture into an animated video through drag operations

Model notes

  • The 1.5 release is based on stable video diffusion
  • The 1.0 release promotes controllable generation across semantic, spatial and temporal aspects
  • Text, image and trajectory are named as the three important control factors

Pros & cons

✓ What we like

  • No video production experience is needed for simple operations
  • Control spans semantic, spatial and temporal aspects
  • Pretrained models can be downloaded for desktop drag generation

! What to watch out for

  • The workflow starts with cloning the repository and downloading pretrained models
  • Results depend on the text, image and trajectory inputs provided

FAQ

What are the three control factors?

The material names text, images and trajectories for the earlier release.

Is video production experience required?

The card states no video production experience is needed for simple operations.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review