TextBox1

The simplest node in the pack, and going by its click-through the one people actually land on. TextBox1 is a multiline text field that outputs a string. You type text in, the same text comes out the other side. That's it - the node's own description is literally "Multiline textbox."

So why does it exist when ComfyUI has text primitives already? Because a bare string node is genuinely useful plumbing. You hold a prompt, a style tag, a chunk of negative text, or any reusable snippet in one place, then wire that string into a CLIP Text Encode - or into several encoders at once, so one edit updates every place the text is used. It's the difference between retyping the same style string into four nodes and editing it once.

How it works

There's no processing. The widget stores whatever you type and passes it through as a STRING. It's a source node: nothing feeds into it, and its value is the text on screen. Because the output is a standard string, it drops into any input that takes text - prompt encoders, string-combiner nodes, filename builders, whatever.

The inputs and outputs that matter

There's exactly one of each:

• text1 (STRING, multiline) - the text you type. Multiline means you can paste a whole prompt with line breaks, not just a single line.

The output is text1 (STRING) - wire it into a CLIP Text Encode's text input, or fan it out to multiple consumers.

Installing it

ComfyUI Manager: search RES4LYF, install, restart. Or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/ClownsharkBatwing/RES4LYF/
cd RES4LYF
pip install -r requirements.txt

pip install in your venv or portable Python, restart, hard-refresh (F5). Nothing to download.

Where people get tripped up

Honestly, there's not much to break - it's a string. The only real question is why you'd install a whole sampler pack for a text box, and the answer is you wouldn't: this comes along for free when you install RES4LYF for the samplers, and it's handy to know it's there. If you're not already running the pack, any of the dozen text nodes from other packs does the identical job.

One small note: it outputs plain text with no weighting or prompt syntax of its own. If you want emphasis, LoRA triggers, or scheduling, that's the job of the encoder or a prompt node downstream - TextBox1 just holds the raw string and hands it over untouched.

Where it does quietly pay off is as a single source of truth. Wire one TextBox1 into both a positive encoder and, say, a filename builder, and your prompt and your saved filename can never drift apart, because they read the same string. Same trick for a shared style suffix fed into several encoders at once: edit it in one place, and every generation in the graph updates. That's the entire value proposition - not the text box itself, but having exactly one copy of the text.