Text

Sometimes the most useful node in a pack is the least impressive one. SimpleText is exactly what it says: a text-in, string-out box. You type (or paste) text, it passes it through. That's the whole feature. And it exists because when you're building a workflow out of the pack's LLM and VLM nodes, you need a dependable place to put your prompt text before it goes somewhere that consumes a STRING.

What it's for

ComfyUI lets you type directly into most text widgets, but not all connections want a literal. When a node takes a STRING input that you want to feed from a central, editable spot - a prompt you reuse across nodes, a system message you tweak between runs, a keyword list you want to keep as a plain variable - SimpleText gives you that single source of truth. It's the pack's own version of the string-input node every graph ends up needing, and it supports dynamic prompts, so you can drop in the {} template syntax.

How it works

One input, one output, no network, no model, no llama.cpp - it's pure Python string plumbing. That's also why it can't really fail: there's no backend to break. The input_text field is multiline, so paragraphs or multi-line prompts work fine.

Inputs and output

• input_text - your text. Multiline.

Output: a STRING with exactly what you put in. Wire it into LLMPromptGenerator, a CLIP Text Encode, or anything else in the pack that eats a string.

Install

Same pack, standard route:

cd ComfyUI/custom_nodes
git clone https://github.com/gokayfem/ComfyUI_VLM_nodes
python -m pip install -r ComfyUI/custom_nodes/ComfyUI_VLM_nodes/requirements.txt

Or ComfyUI Manager → search VLM_nodes.

The honest take

Don't overthink it. If you already have a preferred text node from another pack, you don't need this one - a string is a string, and they're all compatible. But if you're keeping a workflow self-contained to VLM Nodes, this is the clean, dependency-free way to hold your text, and it saves you from typing the same prompt into five different widgets.