Skip to content

ComfyUI-segment-anything-2

Image segmentation through ComfyUI nodes

Visit official site

What ComfyUI-segment-anything-2 is

The library is described as a segment-anything-2 based image segmentation library wired into ComfyUI nodes, usable while still under development.

What you can do with it

  • Segment regions of a still image
  • Exclude areas with negative selection
  • Run segmentation in video mode

What to watch out for

  • Development is still in progress
  • Models are obtained automatically at first use
  • CUDA extension settings can be configured

Pros & cons

✓ What we like

  • CUDA extension support is listed for speed
  • Example code is provided for a quick start

! What to watch out for

  • The library is described as still in development
  • Model files are fetched on first use

FAQ

Does it work offline?

The material states models load offline without a network connection.

Is video supported?

A video mode for dynamic image segmentation is listed.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review