Save Video

The simplest node in the pack: hand it a video, it writes the file to disk. No outputs, no options beyond the one input, nothing to configure. That plainness is deliberate rather than a missing feature - it's the deliberate endpoint of every video pipeline this pack builds.

Why it matters that it's this basic

This pack's video nodes have no in-graph preview - no scrubber, no thumbnail widget, nothing like the inline video player you might be used to from other packs' video-combine nodes. The README says as much directly: it doesn't support previewing video yet, and everything gets written to disk instead. That makes Save Video the actual finish line for any Faceless pipeline, whatever combination of swap, restore, background removal, or merge you've strung together upstream - there's no in-between checkpoint to look at except by saving.

It's one of three ways a Faceless pipeline can end: this node (local disk), Upload Video (a cloud bucket), or Merge Videos (which also writes a result while combining two clips). For anyone just running the pack locally and wanting their finished clip somewhere they can open it, this is the one you want - no credentials, no external account, no configuration at all.

Inputs and outputs

• video - a FACELESS_VIDEO, from anywhere upstream in a Faceless pipeline.

No outputs. This is a terminal node - ComfyUI runs it as a pure side effect, and there's nothing downstream to wire it into.

Installing it

Manager: search Faceless Node for ComfyUI. By hand:

cd ComfyUI/custom_nodes
git clone https://github.com/jeffy5/comfyui-faceless-node
pip install -r requirements.txt

No model needed for this node specifically - you only need the rest of the pack's models set up if whatever's feeding video requires them (a swap or restoration step earlier in the graph, for instance).

Where people get burned

Looking in the wrong output folder. Per the README, video output lands in output/faceless, not ComfyUI's usual output/ root. If a save seems to have vanished, check there first.

Expecting to see something in the graph. There's no preview here and no filename or path input exposed either - this node just writes to the fixed location above. If you want to sanity-check what you're about to save before committing to a full render, run the same FACELESS_VIDEO through a Load Images (Video) node into a PreviewImage in parallel - cheap way to eyeball a frame or two without waiting on the save.

No control over output format or naming. Unlike some video nodes in the wider ComfyUI ecosystem, there's nothing here for codec, filename pattern, or frame rate - this node does one job with zero configuration. If you need that kind of control, it isn't exposed anywhere in this pack.