β¨οΈ CR Load Image List
Batch-load a folder of images for processing
- IMAGE
- show_help
CR Load Image List points at a folder and loads its images as a list - which, in ComfyUI, means every downstream node runs once per image. That's the whole reason to use it: you've got a directory of pictures and you want to run all of them through the same pipeline (upscale, img2img, background removal, tagging) without dragging each one in by hand. Point, set a count, queue once.
It's the input half of a batch-processing workflow. Pair it with an upscaler or an img2img graph and you've turned a one-image tool into a folder cruncher.
How it works
You pick a folder and it reads the image files inside, emitting them as a list. start_index and max_images let you page through a big folder - start at image 20, load the next 10 - which is how you process a large set in manageable chunks or skip ones you've already done. The output being a list is what triggers per-image execution downstream; each image flows through your graph as its own run.
There are two ways to point it at a folder: the input_folder dropdown (which lists subfolders ComfyUI already knows about, under its input directory) and the optional input_path string, which overrides the dropdown and lets you name any folder you like. For anything outside ComfyUI's default input folder, use input_path.
The inputs and outputs that matter
input_folder- dropdown of known input subfolders. Fine for images you've dropped into ComfyUI's input directory.input_path(optional) - a folder path that overrides the dropdown. This is what you'll use for arbitrary directories.start_index- which image to begin at. Good for skipping or paging.max_images- how many to load. Defaults to 1, so bump it up or you'll only get one image and wonder where the batch went.
Outputs:
IMAGE- the images as a list; wire into whatever processes them and it fans out per image.show_help- wiki link, ignore.
How to install it
Comfyroll Studio (Suzie1 and RockOfFire), a mature utility pack - no models, no heavy dependencies.
- ComfyUI Manager - search Comfyroll Studio, install, restart.
- Manual -
cd ComfyUI/custom_nodes && git clone https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes.git, then restart.
Also on CivitAI.
Common issues
The number-one gotcha is max_images defaulting to 1 - plenty of people wire this up, get a single image, and assume it's broken. It isn't; raise max_images to cover the folder. Second is the path: if input_folder's dropdown doesn't show your directory, use input_path with a full path the ComfyUI server can reach. On a hosted or serverless instance that means a path the server sees, not a folder on your own desktop - local paths won't resolve remotely.
And if the images come through in a surprising order, that's filesystem ordering; name your files with zero-padded numbers (001, 002) if sequence matters.
If the node's missing entirely, it's the pack load failure: Comfyroll Studio: Failed to load ... and NameError: name 'CR_HalftoneGrid' is not defined in the log. The NameError is a symptom of a real import error (usually an incompatible Pillow version from another custom node) that can take the whole pack down. Reinstall Pillow (pip install --upgrade --force-reinstall pillow) and restart.
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| input_folder | COMBO | 1 options: 3d | |
| start_index | INT | 00β9999 | β |
| max_images | INT | 11β9999 | β |
| input_pathopt | STRING | β |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| IMAGE | IMAGE | β |
| show_help | STRING | β |