Bria's Ad Delayer API turns a flat ad image into editable layers — a background plate, product and logo cutouts, live text with typography, and vector shapes — returned as one structured JSON document. Trained exclusively on licensed data for safe commercial use.
A single call decomposes a finished ad into its editable parts:
- Background plate: The scene behind everything else, delivered as a hosted image.
- Image cutouts: Products, logos, and other raster elements as transparent PNGs.
- Live text, twice over: Every copy layer comes back in both forms — editable text with full typography (font family, weight, size, letter spacing, line height, alignment) and an SVG of the original letterforms traced from the ad. Edit the copy, or keep the ad exactly as designed; a request parameter picks which renders by default.
- Vector shapes: Boxes with solid or gradient fills, corner radii, and shadows.
Every result is one structured JSON document: a canvas matching the input image, the Google Fonts stylesheets referenced by text layers, and a z-ordered list of layers, each with its bounding box and type-specific styling. Sorting by z_order and absolutely positioning each layer reconstructs the ad — deterministically, and ready for programmatic editing.
The full field-by-field contract, a real layer example, and a rendering guide are in the Ad Delayer API Reference.
The endpoint processes requests asynchronously. The API will immediately return a request_id and a status_url instead of the final result. You must poll the Status Service to track progress; on completion the status body carries a pointer (result.url) to the hosted layer document.
See the full guide at Status Service Documentation for complete details and usage examples.
Inputs are capped at 800 px per dimension on this endpoint. Removing the cap is available to enterprise accounts only — enterprise requests return full-resolution layers. Contact us about enterprise access.