StringToImageAndShow，big title

This is the "FatLabels" node, and it does exactly one homely thing well: it turns text into an image, at a font size you choose. Type your words, pick how big, and out comes a picture of that text. The original use case, straight from the author, is labelling - big readable headings you drop into a workflow so a screen capture or a tutorial screenshot actually reads at a glance, instead of squinting at tiny node titles. And of course you can composite the text image into your generations too.

It's genuinely simple - the author's README calls it their first custom node ever, and it wears that honestly. No frills, just fat text. Which is the appeal.

How it works

You give it a string and a font size. It renders that text onto an image canvas and hands you back the image. Because it's an output node, it'll show you the result on the node face and save it to your ComfyUI output directory, so it doubles as a save-image step for the label. There's an optional font path if you want something other than the default typeface.

The workflow the README describes is about as direct as it gets: search for the node, type your text, choose a font size, wire it to an image preview or save node, and generate. That's the node.

The inputs and outputs that matter

• text (multiline) - what you want rendered. Multi-line, so a heading and a subheading both work.
• font_size (default 36, minimum 1) - how big. This is the whole "fat" in FatLabels - crank it up for a title that dominates a screenshot, keep it modest for an inline caption.
• font_path (optional) - a path to a font file if you want a specific typeface. Leave it empty for the default. The author notes fancier font and colour options were a "maybe someday" - so treat this as the one lever you have over appearance.

Two outputs: images (an IMAGE) - the rendered text, ready to wire into a Preview Image, a Save Image, or a compositing node - and a STRING pass-through of the text.

Installing it

ComfyUI Manager: search Clh Tool for ComfyUI (the README also says you can just search "Fat" or "Labels" once it's installed to find the node), install, restart. Or by hand:

cd ComfyUI/custom_nodes
git clone https://github.com/clhui/ComfyUi-clh-Tool

then restart ComfyUI. If a manual clone doesn't surface the node, check the startup console - rendering text to an image needs an imaging library, and if that import is missing the console will say so; Manager installs a pack's dependencies for you, a bare clone doesn't.

Common issues

The most likely snag is the font. If you point font_path at a file that doesn't exist or isn't a usable font, rendering fails or falls back oddly - so either leave it blank for the default or give it a real path to a .ttf/.otf file. When in doubt, start with no custom font and confirm the plain version works.

Second, mind your canvas versus text size: a huge font_size with a lot of text can overrun or clip depending on how the node sizes its canvas. If a long label gets cut off, shorten the text or drop the size and re-run. And note the near-identical twin in this pack, String2Image clh - same inputs, same "text becomes an image" job. They're effectively the same tool under two names; use whichever you land on and don't overthink the difference.