Image Mixer🧸

Don't let the name fool you: Image Mixer doesn't blend or composite images. It shuffles them. You feed in up to five images, and it randomly reassigns them across five outputs, so the image that went into slot 1 might come out of output 3. It's a randomized reorderer, not a compositor. If you came here looking for a node that merges two pictures together, this isn't it (you want a blend or a latent-composite node for that).

It's a small, specific utility for introducing randomness into which image lands where in a workflow.

How it works

You wire images into Image_1 through Image_5 (any number of them), and each run the node randomly permutes their order and sends them out Output_1 through Output_5. The seed drives the shuffle: randomize it for a fresh order every queue, fix it to reproduce a particular arrangement. That's the whole mechanism. Same images, scrambled positions.

The inputs

• Image_1 ... Image_5 (all optional, IMAGE): the images to shuffle. Connect as many as you have.
• seed: controls the random ordering. Randomize for variety, fix to lock a result you liked.

Outputs are Output_1 ... Output_5, carrying your images in a randomized order.

When you'd use it

This is a "make my batch less predictable" node. Say you've got a fixed set of reference images feeding different branches, and you want their assignment to vary run to run. Image Mixer scrambles which image drives which branch. It's the sibling of Random Image Output in this pack: that one takes a single image and picks one random branch to send it down, while Image Mixer takes several images and reshuffles all of them at once. Both are for controlled chaos in a graph, not for everyday image processing. The pack's How_to_use_RandomizedUtilityNodes.png example shows them in context.

Installing it

Ships with Plush-for-ComfyUI. Install through ComfyUI Manager (search "Plush", install Plush-for-ComfyUI, restart), or clone it manually: cd ComfyUI/custom_nodes && git clone https://github.com/glibsonoran/Plush-for-ComfyUI.git, then pip install -r requirements.txt, restart.

Common issues

The single biggest confusion is the name, so to be clear one more time: it reorders images, it does not merge them. If you expected a blended picture and got one of your originals unchanged, the node is behaving correctly, you just wanted a different node. Beyond that, connect only as many outputs as you have images so a branch doesn't sit empty, and use a fixed seed when you need the arrangement to be repeatable.