Simple LLM: Text Input

Every LLM workflow needs raw text from somewhere - a prompt, a context block, a fact to store. Simple LLM: Text Input is the unglamorous source node for exactly that: a multiline text box that hands its contents out as a STRING. It's the first node in the SimpleLLM family menu and about as simple as a node gets, but it's the thing you'll wire into a surprising number of places.

It belongs to the SimpleLLM set inside ComfyUI-NS-Util, NakamuraShippo's utility pack - an actively-developed collection whose LLM section the README flags as "Implementing the tests now." Despite the name, don't reach for this thinking it does LLM things. It's a string source, nothing more: type text, get text out. Its real value is as the standard input widget for the family, so your workflows don't rely on half a dozen different packs' text nodes.

How it works

The node has a single multiline text field. When the graph runs, it returns that string as its output. There's no transformation, no default content beyond an empty box, no processing - it's a pass-through of whatever you typed. The multiline input is what makes it useful: paste a long context document, a character bio, or a prompt template, and it flows cleanly out the STRING output into whatever expects it.

The inputs that matter

• text - the only input. A multiline string, empty by default. Type or paste your content here.

Output: STRING - the exact text you entered. Wire it into Simple LLM: Merge Text, Simple LLM: Add Memory, or any of the config-driven nodes' prompt paths.

Installing the pack

# ComfyUI Manager → Install via Git URL:
https://github.com/NakamuraShippo/ComfyUI-NS-Util

# or:
cd ComfyUI/custom_nodes
git clone https://github.com/NakamuraShippo/ComfyUI-NS-Util
pip install pyyaml watchdog

Restart ComfyUI. No model files, no dependencies beyond the pack, no network.

Common issues

Honestly, this node is hard to break. The main stumble is expecting it to do something - it doesn't, and that's the point; if you need to edit a string mid-flow, use Simple LLM: Merge Text or the viewer nodes. One practical note: text typed here lives in the workflow file, so keep API keys and anything private out of it (they'll be embedded in saved workflow JSON and shared images). For secrets, use the env-var path on the config nodes instead.