What it does
ComfyTV is an open-source plugin that turns ComfyUI into a canvas for images, video, audio, music, panorama, layers, and 3D. Each stage is its own node with a Run button, and later stages use the last result instead of rerunning the whole graph. The project says about 190 stages ship, including a storyboard, a Director timeline, and an MCP server. Generation uses models you install locally.
Where it fits
Clone it into the ComfyUI you actually run, restart, and generate one image stage before you chain a video stage.
Before you commit
The license is MIT. It is free and runs inside ComfyUI, so you need a machine that can load the models. The GitHub repository was created on 4 June 2026. This check did not install the plugin.
Release & update notes
Dates below refer to the linked announcement or Product Hunt launch. They are separate from the date this profile was added.
- Listed on AI工具集 · ↗
AI工具集 showed ComfyTV as a new listing on 10 November 2026. The repository itself dates from 4 June 2026.
Sources & editorial notes
This profile summarizes public product information and adds workflow guidance. It is not a hands-on performance review. Check the official website for current prices, limits, and availability.
- Official product website
- AI工具集 latest projects
- Official docs: a canvas on ComfyUI with about 190 stages across image, video, audio, music, panorama, layers, and 3D.
- Official repository: MIT license, created 4 June 2026. Install by cloning into ComfyUI custom_nodes.
Source check:
