Nodes/ComfyUI-multiple-inputs/Multiple Text Inputs
ComfyUI Node

Multiple Text Inputs

Concatenate Text From Anywhere, With One Delimiter

By judian17·Created 9 months ago·Updated 9 months ago· 7
Multiple Text Inputs
    • string
    inputcount3
    string_1
    delimiter
    string_2
    string_3

    ComfyUI's prompt boxes are single strings, but your prompt isn't one string - it's a subject plus a style plus a lighting setup plus whatever wildcard segments you're pulling in from other nodes. Multiple Text Inputs is the glue: wire several text sources into it, set one delimiter, and get back a single combined string you can feed straight into a CLIP Text Encode or positive-prompt slot.

    It's the simplest node in this pack by a mile, and honestly the least exciting. But for prompt composition it removes the most annoying kind of busywork: manually retyping the same joined prompt in three different places when you change one style segment.

    How it works

    Mechanically it's a delimiter.join() over the strings you connect, with one friendly detail: empty inputs are skipped. Leave string_3 unconnected and it doesn't produce a dangling comma - the join just skips it. That's the kind of polish that makes the node actually usable in real workflows instead of a footgun.

    A couple of things to know from the schema:

    • string_1 is required and always included - it's the anchor of your composition.
    • The string fields are connection inputs (they carry forceInput in the code), so this node is designed to be fed from other text nodes - wildcard pickers, text routers, seeded-prompt helpers - rather than typed into. Connect whatever upstream text you're composing.

    The inputs that matter

    • inputcount - up to 1000 inputs (you will never need that many). Change it, click "Update inputs", done.
    • string_1 through string_N - the pieces you're joining.
    • delimiter - the separator, and the one real decision you make. Default is empty, which concatenates with no spacing at all; for prompts you almost always want ", " so your parts don't fuse into nonsense like "portraitgoldenhour".

    The single output is string - the joined result. Wire it to a text encoder or a Show Text node to check your work.

    Installing it

    It ships with the rest of the pack, judian17/ComfyUI-multiple-inputs:

    cd ComfyUI/custom_nodes
    git clone https://github.com/judian17/ComfyUI-multiple-inputs
    cd ComfyUI-multiple-inputs
    pip install -r requirements.txt
    

    Or search "ComfyUI-multiple-inputs" in ComfyUI Manager, then restart. Dependencies are just numpy and scipy, no models. All four nodes sit under MyTemplate.

    The one thing to remember

    The delimiter is dumb - it does not add smart spacing or think about grammar. If you want "subject, style, lighting" you type ", " as the delimiter, comma-and-space, yourself. That's the entire learning curve.

    Is this node a revelation? No. It's a small convenience, and the same job can be approximated by chaining ComfyUI's built-in text nodes. But the empty-skip behavior and the dynamic port count make it the tidiest way to compose a prompt from parts that change independently. For anyone doing A/B prompt testing with a fixed base, it's worth the thirty-second install it takes to try.

    CategoryMyTemplate

    Inputs (5)

    NameTypeDefaultDescription
    inputcountINT32–1000
    string_1STRING
    delimiterSTRING
    string_2optSTRING
    string_3optSTRING

    Outputs (1)

    NameTypeDescription
    stringSTRING