Text Overlay Pixaroma

Text Overlay Pixaroma puts a real caption on your image - font, size, rotation, color, an optional background bar - the way you'd do it in a graphics app, except the caption is a node in your workflow and re-renders every time you run. Wire an image in, type your text, and the styled result comes out the other side ready for a Save node.

This is the node for the "I want to add a title to this generated image" workflow that otherwise means exporting, opening GIMP or Photoshop or Canva, typing, and re-importing. In ComfyUI terms it's the text half of the Pixaroma overlay pair: Text Overlay is the full-featured editor for one caption, while its sibling Text Watermark is the no-editor batch stamper. Overlay is the one you want when the text actually matters and you want to place it by hand.

How it works

At render time the node draws the text on top of the incoming image and passes the result out its image output. The styling controls all live on the node body: font (10 bundled, from Inter and Roboto to Oswald, Playfair Display, and Bebas Neue), weight, italic, alignment, line height, letter spacing, opacity, rotation, X/Y position, text color, and an optional background bar. Number fields even accept math expressions like 100+12.

The real selling point is the Open Text Editor button: a fullscreen canvas where you drag the text to move it, drag corners to resize, and drag a round handle on top to rotate - with snap guides to canvas center, thirds, and edges, plus Fit W / Fit H buttons to fill the image. New nodes auto-center their text on whatever image you wire in, so the first render usually looks right without you touching anything.

The inputs

• image (required) - the image to caption.
• text (optional) - a STRING input. When wired, it overrides the text typed on the node at render time (the textarea greys out so you know not to type there). That's the hook for feeding the caption from a prompt, a filename, or any other text source.

One output: image, the input with your text drawn on top. Chain it straight into a Save Image or Preview node.

Install and gotchas

Standard Pixaroma install: ComfyUI Manager (search "Pixaroma") or

cd ComfyUI/custom_nodes
git clone https://github.com/pixaroma/ComfyUI-Pixaroma

Restart. No dependencies or model downloads for this node. (Want custom fonts? The pack reads ComfyUI/models/fonts/ - drop in .ttf/.otf files and they show up in the font picker.)

A couple of practical notes. If your text looks tiny or giant on the first run, it's almost certainly because the node hadn't seen your real image dimensions yet - run once and the auto-center kicks in, then tweak. And the standing Pixaroma advice: after updating the pack, hard-refresh with Ctrl+Shift+R (Cmd+Shift+R on Mac) if a node looks blank or broken - nine times out of ten it's the browser cache, not the code.