Skip to main content
ComfyTV turns ComfyUI into a canvas app for making media — generate, pick, edit, composite, and export across image, video, audio, music, panorama, 2D layers, and 3D, with ~190 stages. Every operation is its own node; results flow downstream automatically. The ComfyTV canvas

​
Install

Restart ComfyUI. Open the Add Node menu (double-click the canvas or right-click) and look for the ComfyTV category. It’s split into sub-menus: Project, Input, Generate, Image, Panorama, Video, VideoFX, Keying, Compose, Timeline, Audio, AudioFX, Music, 3D, Material, Storyboard, Bridge. Add-node menu

​
The basics

​
Each node has its own Run button

Most stages have a ▶ Run button inside their panel. Clicking it runs just that stage — it does not re-run the whole graph. The result appears in the node’s output preview.
Many stages have no Run button at all — crop, rotate, mirror, grid split, panorama viewports, the layer editor, piano rolls, and most FX previews work entirely in your browser and update instantly as you adjust them. A Run (or a final render like FX Chain) is only needed when a file has to be produced.

​
Results flow downstream as snapshots

Once an upstream stage runs, its result is saved as a snapshot; downstream stages use that snapshot directly when they run, without re-running the upstream.

​
Project selector

Drop a Project node to name/switch the current project. Everything you generate is filed under it — with full output history — and restored when you reload the workflow.

​
The sidebar

The ComfyTV sidebar on the left has seven tabs: Workflow (config editor for the selected stage), Assets (media library), Entries (reusable prompt fragments, referenced with @label), Stages (workflows + custom parameters per stage kind, ★ default), Presets, Resources (LUTs / fonts / SoundFonts), and Servers (run stages on other machines). Full tour: sidebar.md.

​
Quick run

  1. Add Generate → Image Stage.
  2. Type a prompt in the node’s text box (e.g. a red apple on a wooden table).
  3. Pick a model in the workflow dropdown — start with Local SD1.5.
  4. Click ▶ Run.
Image stage running The Image Stage produces a set of images (a workflow can output more than one). The first time you Run it, ComfyTV auto-adds an Image Picker wired to its output.

​
Pick a result

Click a thumbnail on the Image Stage itself, or on the auto-spawned Image Picker. Either way an action toolbar (✏️ Edit, 🌐 Panorama, 📐 Multiangle, …) appears next to the picked frame. Image picker with toolbar

​
Zoom into a preview

Hover any output image and use the mouse wheel to zoom (1×–6×), drag to pan, and double-click to reset. Works on every image output preview.

​
Where to next

  • sidebar.md — the 7-tab sidebar and @ prompt references
  • generate.md — generators (text / image / video / music / speech / 3D model)
  • image-tools.md — crop, rotate, mirror, inpaint, erase, cutout, upscale, outpaint, grid split, variations, multi-angle, relight
  • video-and-audio.md — the video & audio suites (~100 video nodes, 30+ audio nodes)
  • making-music.md — scores, piano rolls, SoundFont synthesis
  • compose.md — pickers, A/B compare, storyboard, timeline
  • custom-workflows.md — 🔗 Link workflow (recommended), upload, and the library folder
The full documentation — including a reference page for every node — lives at comfytv.org.