Make Image Batch

Make Image Batch takes a handful of separate images - up to six - and stacks them into a single batch tensor. It's the node you reach for when you've got images arriving from different places in your graph (a couple of loaded reference images, a couple of generated ones) and you need them as one batch rather than as separate wires, because a lot of downstream nodes in this pack and elsewhere only operate on a batch as a unit: Prune By Mask filters a batch, Mask To Region's batch_behavior only means something with more than one image at once, and plenty of samplers and postprocessing nodes expect a batch dimension to work with rather than individually-wired images.

It's a plain utility node - there's no clever masking or blending logic here, just combining separate image inputs into one batch output. The plainness is the point: it's a small, obvious plumbing piece rather than anything you need to reason hard about.

The inputs that matter

• image1 - required, the first (and if it's all you have, the only) image in the batch.
• image2 through image6 - optional. Wire in as many as you actually have; leave the rest unconnected. You're not required to fill all six.

The output is a single IMAGE batch containing every image you wired in, in the order you connected them (image1 first, then image2, and so on).

Since this node isn't covered in the pack's own README - it's newer or simply wasn't documented there - there's no author guidance beyond what the inputs themselves tell you. The safe assumption, and the one that holds for basically every batching node in ComfyUI, is that all the images you feed in need to share the same resolution; batch operations stack tensors along a new dimension, and mismatched width/height between inputs is the standard way a "combine images" node like this one fails. If your source images come from different places in the graph at different sizes, resize them to match before this node rather than after.

Installing it

Make Image Batch is part of Masquerade Nodes. Install via ComfyUI Manager (search "Masquerade Nodes") or manually: cd ComfyUI/custom_nodes && git clone https://github.com/BadCafeCode/masquerade-nodes-comfyui, then restart ComfyUI. No models, no extra pip packages - it's a plumbing node, and installing the pack is all that's needed.

Common issues

If the node errors on run, resolution mismatch between the images you've wired in is the first thing to check - make each input the same width and height before this node, using whatever resize node fits your workflow. Beyond that, remember only image1 is required; if you've got fewer than six images, just leave the unused optional slots disconnected rather than trying to wire in something empty - an unconnected optional input is the expected, working state here, not an error condition.