🔄 Any List

You've got several separate things - three prompts, five images, a handful of latents - and you want them as a single list so a downstream node can iterate over all of them. That's this node. Wire your items into its slots and it hands back one list containing them, in order. The pack sums it up as converting "simple data items into a list," and that's exactly the shape of it.

It's the front end of the pack's list workflow: build a list here, then run it through Filter List, pluck items back out with Any From List, or feed it to any node that processes a list one item at a time.

How it works

ComfyUI treats a list as a sequence of separate items that the graph iterates over, as opposed to a batched tensor processed all together. Any List assembles that list. It has ten optional input slots (any1 through any10), collects whichever ones you've actually connected, and outputs them as one ordered list. Empty slots are simply ignored, so you don't have to fill all ten.

The inputs and outputs that matter

• any1-any10 (all optional, any type) - the items to bundle. Connect as many or as few as you need; order follows the slot numbers.

The single output is any, a list. Wire it into whatever's built to consume a list.

How to install it

Comes with the SDVN pack. ComfyUI Manager, search SDVN_Comfy_node; or:

cd ComfyUI/custom_nodes
git clone https://github.com/StableDiffusionVN/SDVN_Comfy_node
pip install -r custom_nodes/SDVN_Comfy_node/requirements.txt

from the ComfyUI root, then restart. No models or dependencies - it's pure plumbing.

Common issues & troubleshooting

Mixed types in one list. Because the slots accept anything, you can put a string, an int, and an image into the same list. Whether that's fine depends entirely on what reads the list next - a node expecting all images will choke on a stray string. Keep a list homogeneous unless the downstream node genuinely handles mixed content.

List versus batch. This makes a ComfyUI list, not a batched tensor. If a downstream node specifically wants a batch (a single stacked tensor of images, say), a list isn't the same thing and you may need a "make batch" node instead. Most of the pack's own list nodes expect exactly what this produces.

Only connected slots count. If you thought you wired something in but it's not showing up in the list, check the connection actually landed - an unconnected slot contributes nothing, silently.

Order matters. The list comes out in slot order (any1 first). If a downstream index or filter is picking the "wrong" item, check which slot you plugged it into.