Save Image w/o Metadata

By default, ComfyUI's built-in Save Image node bakes your entire workflow - every node, every prompt, every model filename, every seed - straight into the PNG as metadata. Drag that image into anyone's ComfyUI window and they get your whole graph back, no questions asked. That's a great feature when you're sharing a workflow on purpose. It's a bad surprise when you post a finished image somewhere and didn't realize you just handed out your prompt, your LoRA stack, and your custom node list along with it. This node is the fix: a drop-in replacement for Save Image with one extra switch that decides whether that metadata rides along or gets stripped out.

What it is and why you'd reach for it

It's a small, deliberate patch on ComfyUI's own save node - same job, one added control. You still get a filename prefix, you still get PNGs landing in your output folder, but now you also decide per-save whether the workflow gets embedded. That matters more than it sounds like: metadata-scraping is exactly how a lot of "here's the workflow" screenshots on Reddit and Discord actually work - someone downloads your image and loads it straight into their own ComfyUI to see what you did. If you're posting client work, professional renders, or anything you'd rather people not reverse-engineer, that's the exposure this closes.

The inputs that matter

There are four fields and they're all worth knowing:

• images - the image batch coming out of your VAE Decode, same as any save node.
• filename_prefix (default ComfyUI) - the usual naming prefix for output files.
• save_image (default true) - flip this off and the node stops writing files entirely, acting purely as a preview in the UI. Handy for iterating without cluttering your output folder.
• include_workflow (default true) - the whole point of the node. Turn it off and the PNG gets saved with the workflow metadata stripped out.

There's no output socket - this is a terminal node, same as the default Save Image, is_output_node is true and that's the end of the line.

One thing the author flags explicitly: right-clicking an image in the UI to save it manually still respects whatever include_workflow is set to. So it's not just the automatic save path - the toggle governs the image everywhere it leaves the node.

Installing it

Search ComfyUI-Hangover-Nodes in ComfyUI Manager, install, restart. Or by hand:

cd ComfyUI/custom_nodes
git clone https://github.com/Hangover3832/ComfyUI-Hangover-Nodes

Restart ComfyUI and it shows up under the Hangover category. No model downloads, no heavy Python dependencies - it's a pure utility node riding on what ComfyUI core already has, so it should just work the moment it's installed.

Common issues and things to know

Worth being precise about what this actually strips: it's the ComfyUI-specific workflow and prompt PNG text chunks - the JSON blob that reconstructs your graph. It is not a general EXIF/metadata scrubber. If you're worried about other identifying metadata (camera info, GPS, whatever your OS or another tool might attach later), this node alone doesn't cover that - you'd want a proper metadata-stripping pass downstream if that's the bar you need to clear.

The other thing to flag honestly: this repo is marked by its own author as no longer actively maintained, and was only tested on Windows 10. For a node this simple that's low-risk - it's not touching models or GPU memory, just a metadata flag on a PNG write - but if something ever breaks on a fresh ComfyUI core update, don't expect a quick upstream fix. Worst case, dropping include_workflow and just relying on your own file-naming discipline (or a generic strip-metadata node if one's already in your stack) gets you the same privacy outcome.

If you only need this occasionally rather than as your default save node, it's fine to keep both installed side by side - use core Save Image when you want to share workflows, swap to this one when you don't. That's basically what it's for.