Image to List

A lot of ComfyUI nodes take a single image, and a lot of workflows need to hand them several at once - a batch of reference frames, a stack of poses, a few inpainting sources. That's the gap UtilImageToList fills: it collects up to eight IMAGE inputs and packs them into a single IMAGE list that list-aware downstream nodes can iterate over. It's the image twin of the pack's UtilStringToList, and it exists because "give this node N images at once" isn't something core ComfyUI does for you.

How it works

You set a count - 1 to 8 - and the node shows that many image input slots (the dynamic slot show/hide is handled by the pack's JS, so the node grows and shrinks as you change count). Whatever you connect is gathered in order into one output list. Leave a slot unconnected and it's skipped; connect all of them and you get an ordered batch. If somehow nothing ends up connected, it hands back a tiny placeholder image rather than an empty list - a defensive default that keeps the graph from erroring out, though you should treat it as "something's wrong" rather than a feature.

The inputs and output

• count (INT, default 4, 1–8) - how many image slots to show.
• image_1 through image_8 (IMAGE, optional) - the inputs; connect as many as your count allows.
• images (IMAGE list) - the ordered batch output.

That's the whole surface. No resolution matching, no normalization - it's a collector, and it assumes the images are already in the shape your downstream node wants.

Where it fits

The natural partner for this is a list-consuming node: a batch sampler, a per-image loop, or something like the pack's own batch-oriented utilities. It's also handy for feeding a consistent set of references into a ControlNet preprocessor or a character-consistency workflow. For actually loading images from disk into such a list, the pack's UtilLoadImageList is the sibling you want - that one pulls from your input/ folder via dropdowns, this one takes whatever's already in your graph.

Install

Part of the IXIWORKS pack - ComfyUI Manager → search ComfyUI-VideoDescription → Install, restart, or clone the repo into custom_nodes/. No extra dependencies. One caveat: because the image inputs are added dynamically by JavaScript, this node only behaves fully in the web UI - if you build a workflow by hand-editing the API JSON, the image slots won't appear unless the workflow was saved from the UI with them present.