ComfyUI-outputlists-combiner

Did you know that ComfyUI supports so called output lists which tell downstream nodes to execute multiple times within the same run?

https://github.com/user-attachments/assets/303115d3-7c28-42e8-bb52-d02e7cc1022b

Wait, what?

Yes, I didn't know about it either. Apparently everytime you see this symbol: 𝌠 , it means it's an OutputList and a node fetches the items sequentially. This feature seems very under-utilized but it allows values to be split-up, combined and processed sequentially without weird workarounds like for-loops, increment counters or python scripts with external API calls. This project tries to make good use of output lists and integrate well with other custom nodes without enforcing an opionated paradigm:

Features

• OutputList Combinations: create all possible combinations of multiple lists, e.g. prompt combinations, image size variants
• Formatted String: insert variable placeholders to create custom prompts, e.g. a {animal} with a {colored} hat
• Inspect Combo: connect to a COMBO input (like sampler or scheduler) and retrieve all values as a string list, e.g. sampler+scheduler testing, copy+paste model names
• Delimited Strings and FullList: to add backwards compatibility and integrate well with other custom nodes, e.g. grid annotations

Installation

1. Install ComfyUI.
2. Install ComfyUI-Manager
3. Look up this extension in ComfyUI-Manager. If you are installing manually, clone this repository under ComfyUI/custom_nodes.
4. Restart ComfyUI.

Custom nodes

String OutputList

(workflow included)

Create a OutputList by separating the string in the textfield.

inputs

• separator: the string to split the textfield values.
• delimiter: only used for delimited_str (you probably don't need this).
• values: a textfield with the string which will be separated into the OutputList. note that the string is stripped of whitespace before splitting, and each item is again stripped of whitespaces.

outputs

• value: the values from the list. uses OUTPUT_IS_LIST=True internally and will be processed sequentially by corresponding nodes.
• delimited_str: the list joined together by the delimiter, which is often useful for other nodes (like grid annotations).
• full_list: the whole list with OUTPUT_IS_LIST=False in case some nodes require it (you probably don't need this).
• count: the number of items in the list
• inspect_combo: a dummy output only used to pre-fill the list with values from a COMBO input.

https://github.com/user-attachments/assets/a8aed46e-3d95-4422-a60e-7f62641ef036

Number OutputList

(workflow included)

Create a OutputList by generating a numbers of values in a range. Uses numpy.linspace internally because it works more reliably with floating-point values.

inputs

• start: start value to generate the range from
• stop: end value. if endpoint=True this number will be included in the list.
• num: the number of items in the list (not to be confused with a step).
• endpoint: decides if the stop value should be included or excluded in the items.
• delimiter: only used for delimited_str (you probably don't need this).

outputs All the values from the list use OUTPUT_IS_LIST=True and will be processed sequentially by corresponding nodes.

• int: the value converted to int (rounded down/floored)
• float: the value as a float
• str: the value as a string
• full_list: the whole list with OUTPUT_IS_LIST=False in case some nodes require it (you probably don't need this).
• count: the number of items in the list

OutputList Combinations

(workflow included)

Takes up to 4 OutputLists, computes the Cartesian product and outputs each combination splitted up into their elements (unzip). Empty lists are replaced with units of None and unused inputs will always emit None on the respective output.

Examples

[1, 2, 3] x ["A", "B"] = [(1, "A"), (1, "B"), (2, "A"), (2, "B"), (3, "A"), (3, "B")]
[1, 2] x [] x ["A", B"] = [(1, None, "A"), (1, None, "B"), (2, None, "A"), (2, None, "B")]

outputs

• count: the total number of combinations

Formatted String

(workflow included)

Create string with variable placeholders. Uses python's str.format() internally, see https://docs.python.org/3/library/string.html#format-string-syntax . A commonly used format syntax is {a:.2f} to round off a float to 2 decimals.

outputs

• string: the formatted string with all placeholders replaced with their respective values.

Examples

Simple Example

(workflow included)

Just uses a String OutputList to separate a string and produce 4 images in one run.

Combine prompts

(workflow included)

Combines two String OutputList with a OutputList Combinations and merges them into the prompt with Formatted String. It iterates over all combinations of [cat, dog, rat] x [red, green, blue] (3x3=9)

To debug strings it's recommended to use comfyui-custom-scripts Show Text as it outputs a new line for each emitted item.

Combine samplers and schedulers

(workflow included)

Makes use of inspect_combo to populate the String OutputList (unneeded entries were deleted) and connects to the COMBO inputs samplers and schedulers. It iterates over all combinations of [euler, heunpp2, uni_pc_bh2] x [simple, karras, beta] (3x3=9)

Combine numbers

(workflow included)

Makes use of Number OutputList to generate the number ranges [256, 512, 768] x [768, 512, 256] and connects them to the image width and height to produce image variants in portrait, square and landscape.

Integrating LEv145/images-grid-comfy-plugin

(workflow included)

Makes use of delimited_str to correctly file the annotation labels for the grid.

Note that with OutputLists the images from the KSampler are individual items, i.e. batch_size=1. But the image-grid node expects batches, that's why we need to rebatch them into batch_size=9 with the Rebatch images core node.

To understand batch_sizes use comfyui_essentials Debug Tensor Shape which outputs the tensor shape of the image batch in the console (click the >_ button in the lower left corner).

Tensor Shape Debug after KSampler output:

Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]
Shapes found: [[1, 512, 512, 3]]

Tensor shape required for images-grid:

[[9, 512, 512, 3]]

Rebatching images for subgrids

(workflow included)

Makes use of an additional images-grid to convert a image batch_size=4 into a 2x2 grid and then rebatches them for main-grid.

Tensor Shape Debug after KSampler output:

Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]
Shapes found: [[4, 512, 512, 3]]

Tensor Shape Debug after first images-grid:

Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]
Shapes found: [[1, 1024, 1024, 3]]

Tensor Shape Debug required for second images-grid:

[[9, 1024, 1024, 3]]