⌨️ CR Load Image List Plus

When you want to run a whole folder of images through the same workflow - batch upscale, batch background-removal, batch anything - you need a loader that emits a list, and CR Load Image List Plus is Comfyroll's loaded-up version of that. The "Plus" is the extra outputs: alongside the images and masks, it hands you each file's index, filename, and dimensions, plus a total count. That metadata is what turns a dumb batch into something you can label, route, or gate on.

How it works

It reads images from a folder and outputs them as a ComfyUI list. That word matters: a list means the graph executes once per item - each image flows through your nodes independently, in sequence - as opposed to a batch, which is one stacked tensor processed together. Lists are how you process files of different sizes without forcing them all to match. (When you do need them stacked back into a batch - say, to save as one strip or feed an animation node - that's what CR Batch Images From List is for.)

The inputs and outputs that matter

• input_folder - a dropdown of subfolders inside ComfyUI's input directory.
• input_path (optional STRING) - an arbitrary path if your images live elsewhere; this overrides the dropdown.
• start_index (INT, default 0) - skip the first N files.
• max_images (INT, default 1) - how many to load. Bump this - the default of 1 loads a single image, which surprises people expecting the whole folder.

Outputs (the image/mask/index/filename ones are lists): IMAGE, MASK, index, filename, width, height, list_length, and show_help. Wire IMAGE into your processing chain; use filename to name outputs so a batch stays traceable.

Installing it

Comfyroll Studio (Suzie1 / RockOfFire) is a large, established utility pack. Grab it in ComfyUI Manager (search "Comfyroll Studio," install, restart) or clone it:

cd ComfyUI/custom_nodes
git clone https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes.git

then restart. No models needed.

Common issues

The number-one "it only did one image" complaint is max_images still on its default of 1. Set it to the number of files you want (or something large to grab everything).

Second, list vs batch trips up a lot of people: if a downstream node throws a shape or type error, or you expected a grid and got single images, you're handing a list to something that wanted a batch. Convert with CR Batch Images From List. Also note that a hosted/serverless ComfyUI won't see your local folders - you'll want the images uploaded to the environment's input directory, or use input_path pointing somewhere the runner can actually read. If the whole pack fails to load at startup with a traceback, reinstall it cleanly through Manager.