Save Video MXD
Save a new video version next to the original without clobbering it
- video
If you've ever re-rendered a video edit and then spent five minutes figuring out which of seven similarly-named mp4s is the new one, this node is aimed at you. Save Video MXD is the video output node from Maxed Out's pack, and its whole thing is versioning: save a new version next to the original, with clean counters, instead of overwriting or scattering files.
The description on the node puts it plainly: "Save a new version next to the original with clean counters." In a workflow where you're iterating on a video edit, render after render, that's exactly the behavior you want - each pass becomes name_001, name_002, and so on, sitting alongside the source so you can compare without hunting.
How it works
It's an output node: video in, nothing out. You give it the path to the original video (video_path), and it writes the new version beside it with a counter appended. The other inputs give you control over the output:
save_to_outputs- a toggle (default off) for sending the file to ComfyUI's output directory instead of next to the original.format- the container format, defaulting toauto.codec- the video codec, also defaulting toauto, so it picks something sensible unless you specifically want a particular codec.
Because format and codec both default to auto, a beginner can ignore them entirely - the node figures out a working container from your video.
Installing it
Part of the ComfyUI-MaxedOut pack. ComfyUI Manager search Maxed Out, or:
cd ComfyUI\custom_nodes
git clone https://github.com/Maxed-Out-99/ComfyUI-MaxedOut.git
Restart ComfyUI. No extra dependencies or model downloads.
The honest caveats
This one's worth a "trust but verify": since it's the newest of the MXD video nodes, behavior details like exactly where the counter goes live in the pack's repo and can shift between versions. If a saved file lands somewhere unexpected, check video_path first - it's a required string input, and the node saves relative to what you give it. And like the rest of the pack, it speaks the ComfyUI VIDEO type, so if you're working with frame batches you'll need to convert before it'll accept anything.
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| video | VIDEO | The video to save. | |
| filename_prefix | STRING | video/ComfyUI | The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes. |
| format | COMBO | auto | The format to save the video as. |
| codec | COMBO | auto | The codec to use for the video. |
| embed_workflow | BOOLEAN | true | When high_workflow is connected, merge it into this video's embedded workflow so dragging the final video into ComfyUI shows both the high-noise stage and this stage together. |
| high_workflowopt | STRING | Connect a Load Latent node's 'high_workflow' output here to carry the high-noise stage's workflow into this video's metadata. |
Outputs (0)
No outputs