Nodes/ComfyUI Easy Use/easy wildcardsMatrix
ComfyUI Node Runs on cloud

easy wildcardsMatrix

Sweep every combination of your wildcards, not a random one

By yolain·Created 3 years ago·Updated 8 days ago· 2,633
easy wildcardsMatrix
    • populated_text
    • total
    • factors
    text
    Select to add LoRA
    Select to add Wildcard
    offset0
    output_limit1

    Regular wildcards pick one random option per run. This node does the opposite: it enumerates every combination of the wildcards in your prompt and hands you the whole list. If your text has {red|blue|green} and {cat|dog}, you don't get one lucky roll - you get all six. That's the difference between a slot machine and a contact sheet, and it's the tool you want when you're systematically exploring a prompt space instead of fishing.

    How it works

    You write a prompt with wildcard tokens and LoRA tags in the text box. The node expands the combinatorial product of every option and returns each fully-populated prompt as a separate string in a list. Feed that list into a sampler and ComfyUI runs one generation per combination - a proper matrix sweep. It also tells you how many combinations exist, so you know before you hit go whether you're about to render 6 images or 600.

    This is a real, maintained feature (it arrived in v1.3.0), and it plugs the exact gap people keep hitting when they move to ComfyUI from A1111 - the ability to fire off a structured batch of prompt variations and walk away.

    The inputs and outputs that matter

    • text - your prompt, with wildcard tokens. This is the whole game; write it with the variations you want to sweep.
    • Select to add LoRA and Select to add Wildcard - convenience dropdowns that insert a LoRA tag or a wildcard file reference into the text box for you, so you don't have to remember exact filenames.
    • offset - where in the combination list to start. With a huge matrix you can't render everything at once, so offset lets you page through it in chunks.
    • output_limit (optional, default 1) - how many combinations to emit at a time. Set it to -1 to output all of them; leave it at a small number to preview a few.

    Outputs: populated_text (a list of finished prompt strings - the thing you wire into your text-encode or sampler), total (the integer count of all combinations), and factors (the per-wildcard option counts, useful for sanity-checking that your matrix is the size you expected).

    Installing it

    Grab the pack through ComfyUI Manager - search ComfyUI Easy Use, install, restart. Or clone it:

    cd ComfyUI/custom_nodes
    git clone https://github.com/yolain/ComfyUI-Easy-Use
    

    Install the requirements and restart. Wildcard files are plain .txt lists in the pack's wildcards folder; the dropdown reads from there. For LoRA Block Weight syntax inside your wildcards you'll also want the ComfyUI-Inspire-Pack installed.

    Where it bites

    The obvious one: the matrix explodes. Combinations multiply, so three wildcards with ten options each is a thousand renders, not thirty. Check the total output before you queue anything, and use offset plus output_limit to work through big sweeps in batches rather than trying to render the entire cross-product in one prompt. If you actually wanted random variety rather than exhaustive coverage, this is the wrong node - reach for easy wildcards instead, which rolls one option per run. Matrix is for when you want to see all of them, on purpose.

    CategoryEasyUse/Prompt

    Inputs (5)

    NameTypeDefaultDescription
    textSTRING
    Select to add LoRACOMBO1 options: Select the LoRA to add to the text
    Select to add WildcardCOMBO2 options: Select the Wildcard to add to the text, __example__
    offsetINT00–1125899906842624
    output_limitoptINT1Output All Probilities

    Outputs (3)

    NameTypeDescription
    populated_textSTRING
    totalINT
    factorsINT