Load Video (Yogurt Nodes)
Pull a video in, trim it, feed your video workflow
- start_time
- end_time
- VIDEO
If you're doing video in ComfyUI - LTX, Wan, Hunyuan, whatever the current thing is - the graph starts with a video sitting on disk and ends with one on disk, and in between you need something to bring that clip in. YogurtLoadVideo is that something: pick a file from your input folder, optionally trim the front and back, get a VIDEO output.
It's part of ComfyUI-YogurtNodes, yogurt7771's 150+ node all-in-one pack - obscure, MIT-licensed, auto-generated README, no community footprint - and it's one of those IO nodes that's unremarkable until you realize how often video workflows need exactly a "load with a start/end trim" and how rarely stock ComfyUI ships one with trim built in.
How it works
The video input is a dropdown populated at runtime from your ComfyUI input directory - drop clip.mp4 into ComfyUI/input and it appears. Then two trimming fields, both in seconds: start_time (0 = from the beginning) and end_time.
The usual gotcha applies: end_time of 0 means "no end limit", not "zero seconds". Leave it at 0 to load the whole clip; set 15.0 to keep only the first fifteen seconds. Both step at 0.01s, so you can cut fairly precisely.
The output is a single VIDEO in the standard ComfyUI video format, which wires into whatever video pipeline you're running - video-to-frames, interpolation, VACE-style editing, or a combine node at the end of the line.
Where you'll use it
- Loading a source clip for img2vid, video editing, or extension workflows.
- Trimming out a section you want to work on, so downstream frame extraction doesn't chew through the whole file.
- Feeding a reference clip into a pipeline without preprocessing it elsewhere.
The folder rule is the one habit to keep: it has to live in ComfyUI/input to appear in the dropdown. If your clip lives elsewhere on disk, the pack also ships Load Video Path, which takes an absolute path string instead - same idea, no file-copying.
Install
ComfyUI Manager, search ComfyUI-YogurtNodes, install, restart. Or:
cd ComfyUI/custom_nodes
git clone https://github.com/yogurt7771/ComfyUI-YogurtNodes
cd ComfyUI-YogurtNodes
pip install -r requirements.txt
Restart, look under "Yogurt Nodes". Light dependencies (numpy, pillow, requests, opencv-python), no model downloads - the heavy AI-client libraries in the pack only load when you touch the LLM nodes, so installing for a loader gets you nothing extra to babysit.
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| video | COMBO | 0 options: | |
| start_time | FLOAT,INT | 0 | Start time in seconds. 0 means from the beginning. |
| end_time | FLOAT,INT | 0 | End time in seconds. 0 means no end limit. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| VIDEO | VIDEO | — |