Superside Load Image

Every workflow starts with loading an image, and this is the Superside pack's own version of core ComfyUI's LoadImage - the same dropdown, the same upload button, the same two outputs, deliberately reimplemented so a workflow built entirely from this package doesn't depend on core's node set for its most basic operation. If you've used LoadImage once, you already know how to use this.

Why would a pack ship a clone of the single most basic core node? It's the same logic that runs through this whole pack's utility shelf: a production workflow that must survive being moved, exported, or run on a locked-down backend shouldn't have an invisible dependency on "core nodes behave exactly like this forever." By reimplementing the contract, the pack guarantees the behavior it relies on. For you, the practical difference is nil in day-to-day use - it's a Load Image that lives under the Superside category.

The inputs and outputs

One input: image, a dropdown of every image in ComfyUI's input folder - plus an upload button to drop a new file in from disk. Choose one, done.

Two outputs, and this is the bit people forget about LoadImage-style nodes:

• IMAGE - the image as a tensor, ready to feed anything.
• MASK - the alpha channel, exposed as a mask. If the PNG has transparency, this output gives you a ready-made selection mask of the opaque area. Great for cutouts; easy to miss if you've never scrolled to the second output socket.

The node also respects ComfyUI's workflow-in-the-PNG convention - load a PNG that was saved from ComfyUI and the workflow metadata travels with it (that's the image-io-metadata behavior worth knowing about generally).

Installing it

Standard pack install:

cd ComfyUI/custom_nodes
git clone https://github.com/Superside/comfyui-superside-nodes
pip install -r requirements.txt

Restart ComfyUI and it's under Superside. No API key, no model download, nothing to configure - it's pure local file I/O.

When you'd actually use it over the core node

If you're not building a Superside-only workflow, the core LoadImage is right there and there's no reason to swap. If you're loading this pack's own workflow templates - the ones that assume you have all its utility nodes - then this node is what they use, and knowing it's a faithful clone means you can trust it exactly like the original. Either way: dropdown, upload, IMAGE + MASK out, and the most predictable node in the pack.