Image List to Image Batch

ComfyUI has two ways of carrying multiple images and they are not interchangeable, which trips up basically everyone at some point. A list is many separate items that nodes process one at a time; a batch is a single stacked tensor that nodes handle in one go. Plenty of nodes will only accept one shape or the other, and when your images are in the wrong one you get an error or, worse, only the first image getting processed. Image List to Image Batch converts the list form into the batch form. That's its whole job, and it's the fix for a very specific and very common wiring headache.

How it works

It takes an image list and stacks the items into one batched image tensor. Where a list makes a downstream node fire once per item (or silently take only the first), the batch lets it operate on all of them together - for example feeding several images into a single node that expects a batch, or previewing them as one strip. It's the inverse of Image Batch to Image List, which goes the other direction.

The inputs and outputs that matter

There's really only one input and one output here, which is a nice change of pace:

• images (IMAGE) - the incoming image list. This node is meaningful precisely because the input arrives as a list of separate images.
• Output: a single batched IMAGE you route into whatever node needed a batch.

No settings, no thresholds. You wire it in when the type dance demands it.

How to install it

ComfyUI Manager: search ComfyUI Impact Pack, Install, restart.

Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/ltdrdata/ComfyUI-Impact-Pack
cd ComfyUI-Impact-Pack
pip install -r requirements.txt

Run the pip in ComfyUI's Python environment and restart. No models involved - this is a pure data-shape utility.

Common issues & troubleshooting

Only the first image comes through. That's the classic list-vs-batch symptom, and often the reason you reached for this node. If it persists after this node, check that the node consuming the batch actually iterates the batch dimension rather than grabbing index 0.

Dimension mismatch when batching. A batch is a single stacked tensor, so all images in it have to share the same width and height - that's a property of batches in general, not a quirk of this node. If your list holds images of different sizes, stacking them is ambiguous. When you specifically want to combine differently-sized images into one batch and have them reconciled, reach for Make Image Batch instead, which is built to scale inputs to fit.

"Do I even need this?" test. If the downstream node accepts a list happily, you don't. This node exists only to satisfy nodes that demand a batch. When in doubt, try wiring your list directly first and add this converter only if you hit a type error.

Confused which direction you want? List → batch is this node. Batch → list (to make a node run once per image) is Image Batch to Image List. The naming is literal; read it left to right.