JoinImageBatch

You ran a batch - four seeds, four images - and now you want them as a single picture you can glance at or post, not four separate files. easy joinImageBatch takes a batch of images and lays them out into one combined image, side by side or stacked. Instant contact sheet.

It's an image utility from yolain's ComfyUI-Easy-Use pack (one of the originals, in there since v1.0.4). Simple job, genuinely handy - comparison grids are half of what people do while dialing in a prompt, and this saves you from exporting a pile of PNGs and assembling them somewhere else.

How it works

The input is a whole images batch - the kind a sampler produces when batch_size is more than 1. The node arranges every image in that batch into a single output image. The mode input picks the layout: horizontal puts them in a row, vertical stacks them in a column. It walks the batch and concatenates, so a batch of 4 becomes one 4-wide (or 4-tall) image.

That batch-oriented design is the whole distinction from easy imageConcat. imageConcat joins exactly two images you wire in by hand. joinImageBatch takes an arbitrary batch and lays out all of it at once. If you're combining the output of a batch generation, this is the node; if you're gluing two specific unrelated images, that's imageConcat.

Inputs and outputs

• images - the batch to combine.
• mode - horizontal (a row) or vertical (a column).

Output: a single image, the combined strip. Send it to a preview, a save node, or easy imageConcat if you want to then bolt a caption panel beside it.

Installing it

Comes with the pack. ComfyUI Manager → "ComfyUI Easy Use" → install → restart. Or cd ComfyUI/custom_nodes && git clone https://github.com/yolain/ComfyUI-Easy-Use, run install.bat / pip install -r requirements.txt, restart. First-launch theme swap is the pack's UI - Settings → Color Palette to undo.

Common issues

Two things to keep in mind. First, this expects an actual batch on one wire, not several separate image outputs. If your images are coming from different nodes, you need to combine them into a batch first (an image-batch node), then feed that in - joinImageBatch can't reach out and gather images that aren't already batched together.

Second, the images should share dimensions for a clean result. A batch straight out of one sampler will all be the same size, so that's usually automatic. But if you've assembled a batch from differently sized images, the join can look ragged. And there's a practical ceiling: joining a batch of 30 large images makes one enormous image that'll eat memory and may be unwieldy to save - for big comparison grids, keep the batch modest or downscale first.