Load Video from Asset

Pick a video from the ComfyTV project asset library for clip, demux, upscale, or other downstream steps—reuse in-project media, not raw input/ files.

Load Video from Asset · screenshot coming soon

What this node does

Load Video from Asset offers the same browsing UX as the image asset loader: categories, search, thumbnails/previews. A click immediately outputs a COMFYTV_VIDEO snapshot.

Library videos come from Video Stage runs, timeline renders, sidebar imports, etc.—scoped to the current Project, separate from ComfyUI’s global input/ folder.

For MP4 on disk that never entered ComfyTV, use Load Video (input/) instead.

When to use it

• After Timeline Render or Video Stage, branch for Upscale / Clip.
• Pick a historical version from the library (Compare after Extract Frame if needed).
• Load as asset node from hover menus for a pre-filled node.
• Multi-shot projects: reuse clips without re-running upstream generators.

How it works (why ComfyTV is designed this way)

• No ▶ Run: Click asset → output; instant Input stage.
• Snapshots: Downstream reads URL; won’t re-run the stage that created the video.
• Library vs input/: Same as Load Image from Asset—project-scoped vs global disk folder.
• No workflow: Does not invoke workflows/video/ generators.

Types (COMFYTV_* vs native ComfyUI)

ComfyTV type	What it is	vs ComfyUI native
COMFYTV_VIDEO	Video URL snapshot	Not VIDEO tensor
COMFYTV_IMAGE	Single frame	After Extract Frame, use Compare
Native VIDEO	Tensor	→ ComfyTV Video Bridge

Details: bridges.md

UI and parameters

Asset picker

• Category tabs, search, clickable entries/thumbnails.
• Empty library → run Video Stage or sidebar import first.

asset_url / asset_id / category (hidden)

• UI-maintained; asset_url is the output payload.
• category persists last filter.

project_id / parent_output_id

• Hidden; bound to Project.

Outputs

Output	Type	Meaning	Typical downstream
video	COMFYTV_VIDEO	Selected video snapshot	Video Clip, Crop, Upscale, Demux, Extract Frame

Step by step for beginners

1. Set up a Project node.
2. Run Video Stage or import video into the library.
3. Add Load Video from Asset.
4. Browse and click the target video.
5. Wire to Video Clip or other Video stages.
6. ▶ Run downstream.

Full guides (recommended reading)

This page covers one node only. For end-to-end workflows, multi-stage pipelines, type conversion, and design rationale, see the ComfyTV user guides on GitHub:

Guide	Contents
Getting started	Install, canvas basics, per-stage Run, snapshots, Project, Image Picker
Video and audio	Clip, crop, resize, extract frame, demux vs Generate video
Compose	Image Picker, Compare, Storyboard→Shot Images, timeline

Repository and workflows

Resource	Link
GitHub repository	https://github.com/jtydhr88/ComfyTV
User guides index	https://github.com/jtydhr88/ComfyTV/tree/main/docs
Built-in workflows	https://github.com/jtydhr88/ComfyTV/tree/main/workflows
Model checklist	https://github.com/jtydhr88/ComfyTV/blob/main/docs/models.md
Custom workflows	https://github.com/jtydhr88/ComfyTV/blob/main/docs/custom-workflows.md

FAQ

Q: vs Load Video (input/)?
A: input/ = external raw files. Library = project generated/imported. Generate with Video Stage; reuse with this node.

Q: Missing a video I just ran?
A: Same Project? Refresh the node after Run completes.

Q: Type errors?
A: Downstream needs COMFYTV_VIDEO; native tensors need Bridge.

Q: Replace Video Stage?
A: No—this selects only; it does not generate video.

Related nodes

• Load Video — input/ files.
• Video Stage — writes new videos to the library.
• Timeline Render — timeline export into library.
• Load Image from Asset / Load Audio from Asset — sibling Input nodes.