Image List To Image Batch

The mirror image of easy imageBatchToImageList. This one takes a list of images - separate items ComfyUI was processing one at a time - and packs them into a single batch tensor. You reach for it when you've been iterating per-image and now need everything together in one shot: to feed a sampler, to make a grid, to save as one animated output, to run a node that expects a batched input.

If you don't already know the batch-vs-list distinction, it's the thing that trips up half of ComfyUI at some point: a batch is one stacked tensor processed together; a list is individual items processed separately. Some nodes want one, some want the other, and these two conversion nodes are how you switch between them.

How it works

It stacks the incoming list items along the batch dimension into a single tensor. That's why they have to be the same size (see below) - you can't stack tensors of different shapes.

The inputs and outputs that matter

• images - the list of images to combine.
• IMAGE (output) - one batched IMAGE containing all of them.

No knobs. Plug in a list, get a batch.

How to install it

ComfyUI Manager: search ComfyUI-Easy-Use, install, restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/yolain/ComfyUI-Easy-Use

then install.bat / pip install -r requirements.txt, restart. No models.

Common issues & troubleshooting

Shape mismatch error. The single most common failure: your list items aren't all the same width and height, so they can't be stacked. This happens constantly when the list came from images of different sizes (say, mixed source photos). Normalize them to one size first - easy imageScaleDown with crop: center, or any resize node - then batch.

Nothing seems to change downstream. If the node after this one processes lists and batches identically, you won't see a difference. This conversion matters specifically for nodes that behave differently on a batch - a sampler running everything in one pass, a save-as-grid, an animation assembler.

Order matters for animations. When you're stitching frames, the batch order is the frame order. If a shuffle happened upstream, the batch will carry that order through. Check the sequence before you commit it to a video node.

A one-item list is a no-op. If the "list" only holds a single image, batching it just gives you a batch of one - nothing breaks, but nothing meaningful happened either. This node earns its place when you've genuinely fanned out to per-image processing (captioning each frame, detailing each face) and now need them recombined; on a single image it's just overhead you can delete.

Where the list usually comes from. Nine times out of ten the list on this node's input started life as a batch that easy imageBatchToImageList split, so some node in between could run per-image. This node closes that round trip. If you find yourself splitting to a list and immediately batching again with nothing in between, drop both nodes - you never needed the detour.