ComfyTV Lands A Visual Canvas That Snaps GenAI Into Independent Steps
ComfyTV is a new canvas-based app that integrates natively with ComfyUI, turning the node-based interface into a visual workflow canvas. It lets you chain generative AI tasks like text-to-image, inpainting, and video generation step by step. Each operation runs as its own node rather than through ComfyUI’s global queue, so re-running one stage doesn’t trigger the whole chain.
The project was created by developer Jtydhr88, who built it to feel like a true ComfyUI-native canvas. It reads your existing ComfyUI subgraphs and plugins, and you can drop in any custom workflow JSON file without extra Python coding. The tool is open source and runs entirely on local hardware with no external API calls.
Run stages independently on your own hardware
Key Features
- Each node runs on its own, not globally.
- Downstream stages use last output snapshots.
- Drop in your own ComfyUI workflows.
- Generate images, video, audio, and panoramas.
- Built-in mask painter and 3D camera editors.
- Project saves stages and restores on reload.
Anyone who already works with ComfyUI can benefit from a canvas that visualizes complex pipelines. The tool is ideal for creatives who need precise control over multi-step synthesis, from initial image generation to video editing. Because everything runs locally, you keep full ownership of your data and models.
What the developer says
The developer notes that ComfyTV adds no extra Python dependencies beyond what ComfyUI already uses. It integrates with third-party plugins like IPAdapter through special Bridge nodes. A step-by-step guide covers everything from installation to adding your own workflows without touching code.
"ComfyTV — the canvas-based app that truly belongs to ComfyUI." — Source: GitHub
Project Links
