List your AI tool on AI Orion for free. Submit a tool

Steerable Motion

A ComfyUI node for driving videos using batches of images.

About Steerable Motion

Steerable Motion is a set of ComfyUI custom nodes and workflows for travelling between images to create video. It is built for ComfyUI users, including AI artists, who want fine-grained control over how a sequence of images is animated into motion rather than relying on a general text-to-video model that performs well without extensive manual tuning.

Steerable Motion offers two approaches. The Wan approach uses VACE to create anchor images and continuations chained together from previous images. The Animatediff approach combines IP-Adapter and SparseCtrl to travel between a set of input images, with tweaks to settings like each frame's length of influence changing the resulting motion significantly. The tool behaves more like a paint brush than an out-of-the-box generator, rewarding trial and error to learn which settings and input images work best.

The Animatediff approach ships with five basic workflows, each with its own motion character, ranging from smooth to more blocky or realistic, mostly driven by IP-Adapter configuration and base model choice. The Wan approach provides a sample workflow built on ComfyUI-WanVideoWrapper for chaining VACE-generated anchor images into continuous video. Both approaches require installing the needed custom nodes and models through ComfyUI Manager before a workflow will run.

Steerable Motion's first-party code is released under the Apache License 2.0, while bundled third-party components keep their own licenses, including ComfyUI IPAdapter Plus under GPLv3 and ComfyUI Frame Interpolation under MIT. It runs as a node inside ComfyUI and requires ComfyUI and ComfyUI Manager to be installed first; there is no pricing, as it is a free, open-source project.

Use cases

  • Animating a sequence of images into a continuous video
  • Experimenting with different motion styles, from smooth to realistic to blocky

Key features

  • Generates video by travelling between a batch of input images
  • Uses VACE to create anchor images and chained continuations (Wan approach)
  • Combines IP-Adapter and SparseCtrl to travel between images (Animatediff approach)
  • Ships five preset Animatediff workflows with distinct motion characteristics
  • Lets frame-length tweaks change how strongly each image influences the motion
  • Installs missing dependent nodes and models directly through ComfyUI Manager

Pros and cons

Pros

Open source under the Apache License 2.0
Provides multiple preset workflows so a user can pick a starting point

Cons

Does not perform well out of the box like text-to-video; requires trial and error to tune
Bundled third-party components carry their own licenses that must also be honored

Reviews

Loading reviews…

Alternatives to Steerable Motion

More in AI Video Generator