Input Text

Input Text does exactly one thing: it gives you a big multiline text box that outputs a string. Nothing else. No transformation, no prompt magic, no model. Which sounds pointless until you've spent ten minutes fighting ComfyUI's default single-line string widgets, or you're building one of this pack's API workflows and need a clean place to type a long prompt.

The README's own quick-start mapping lists it as a starter node for text generation: type your prompt here, wire text into an LLM API node, get a response back. That's the whole job. It exists because when you're chaining text through API nodes, you want the text entry point to be obvious and easy to edit - not buried in a widget on some node you only half understand.

How it works

It's a pass-through. The text input (multiline, so you can paste a whole paragraph or a caption) is returned unchanged as the text output. That's the entire mechanism - there's no tokenizing, no truncating, no default value to trip over. What you type is what comes out.

Two details make it slightly more useful than the built-in equivalent:

• dynamicPrompts is enabled on the widget. That means if you have ComfyUI's Dynamic Prompts feature on, the text you type participates in template expansion - {red|blue|green} style randomization works right in the box. Handy for generating prompt variations across a batch.
• It's a normal string widget, so in the ComfyUI graph you can right-click it and convert the widget to an input. That turns the box into a wire point - another node's output can feed it, and it becomes a convenient place to splice or override text mid-graph.

The output is a single STRING, which is exactly what the pack's LLM, vision, and image-gen API nodes expect on their prompt inputs. The HyprLab node in this same pack even marks its prompt as forceInput, so a text box like this is the natural thing to wire into it.

When you'd actually use it

If you're just running a normal local generation workflow, you probably don't need this - ComfyUI's built-in text nodes cover you. Reach for it when:

• You're typing long, multiline prompts for an API text-generation node and want them readable and editable.
• You want a single shared text source feeding multiple nodes (an LLM prompt and a caption, say) so you edit one box instead of five.
• You're building a workflow template for other people and want the "type your text here" spot to be obvious.

Installing it

It ships in the ComfyUI_BillBum_Nodes pack alongside the other nodes covered here. Install once via ComfyUI Manager (search billbum) and you get the whole set:

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

Dependencies are light - openai, requests, pillow, numpy, tenacity, tiktoken, urlextract - no models, no downloads. Restart ComfyUI and the node shows up under string processing.

Troubleshooting

There's almost nothing to break here, which is the point. If the box won't show in your workflow, it means the pack isn't installed or ComfyUI wasn't restarted after install. If you wire it into something and get nothing out, check that you actually typed into this node's box and not a copy elsewhere in the graph. If dynamic prompts don't seem to expand, that's the global Dynamic Prompts feature not being enabled - the node just marks the widget as eligible; it doesn't turn the feature on itself.

Boring, dependable, and occasionally exactly what you needed.