Image Add Multi

Blending two images is a one-node job. Blending five or six the same way normally means chaining that same blend node over and over, feeding each output into the next input - workable, but it clutters your graph fast and makes it annoying to adjust the mix later. ImageAddMulti collapses that chain into a single node with a growable number of image inputs, so you can add, subtract, multiply, or difference a whole stack of images together in one place.

How it works

You set inputcount to however many images you want to combine and click the node's "Update" button - a KJNodes pattern that shows up on several nodes in this pack - and it grows extra image_N sockets to match. Each image gets combined into a running result according to whichever blending mode you picked, weighted by blend_amount, producing one output batch.

The inputs and outputs that matter

• inputcount (default 2, minimum 2, up to 1000) - how many image inputs the node exposes. Bump this and hit "Update" before wiring more images in; the sockets only appear after you update the count, not just by typing a higher number.
• image_1 and image_2 (required) - the first two images; every additional slot beyond these two follows the same naming pattern (image_3, image_4, and so on) once you've grown the input count.
• blending (default add, one of add / subtract / multiply / difference) - the operation applied across all the images. add brightens by summing values, multiply darkens and tends to preserve shadow detail, difference is the classic way to visualize exactly where two images diverge.
• blend_amount (default 0.5, range 0–1) - how strongly the blend operation is applied versus the base image.

Output is a single images batch - the combined result, ready for a preview node or the next step in your pipeline.

How to install it

Via ComfyUI Manager: search "KJNodes for ComfyUI," install, restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/kijai/ComfyUI-KJNodes
pip install -r ComfyUI-KJNodes/requirements.txt

then restart. No extra dependencies - this is straightforward tensor math.

Common issues & troubleshooting

New image sockets don't appear after raising inputcount. Click "Update" - raising the number alone doesn't grow the node; that button is what actually adds the widgets. This dynamic-input pattern trips people up across the pack because it's not obvious from the UI alone.

Result looks blown out or completely black. add mode with several images and a high blend_amount compounds fast - brightness stacks with every additional image, and you can clip to white well before you've combined everything you meant to. Drop blend_amount or switch to a mode like difference or multiply that doesn't compound the same way.

Images need to match in size and batch length. Like most image-combining nodes in ComfyUI, mismatched resolutions or batch sizes across your image_N inputs will error rather than silently resize - resize or crop everything to match before feeding it in here.