Firefly Composite API for product photos vs Sume reference edit
Firefly Composite Operations blend a product photo into a generated scene. On Sume, send the photo as an input reference, with an optional mask_url.
Firefly's Composite Operations APIs (Object Composite, Adaptive Composite and Precise Composite) blend a product shot into a generated scene. Sume has no composite endpoint in its image docs. The closest path is POST /v1/images with the product photo in input_references, a prompt describing the scene, and on ChatGPT Image 2.5 an optional mask_url.
Adobe facts are from its API overview; Sume facts from the Image API docs, read 2026-10-01.
What do the Firefly composite APIs offer?
Adobe describes uploading an image such as a product photo and using a text prompt to generate a composited scene. Precise Composite adds harmonization strength from 0 to 100 percent for lighting and shadows. Adaptive Composite can place products into existing or custom-generated backgrounds and adjusts object rotation to the scene.
What does the Sume path look like?
Both ChatGPT Image 2.5 variants support text-to-image, up to 16 image references and an optional mask_url. The docs describe mask_url as an optional public HTTPS mask URL for ChatGPT Image 2.5 edits. Reference URLs must be public HTTPS.
| Control | Firefly composites | Sume images |
|---|---|---|
| Product input | Uploaded product image | input_references (up to 16 on ChatGPT Image 2.5) |
| Scene | Text prompt | Text prompt |
| Limit the edit area | Not covered on the page read | Optional mask_url |
| Lighting harmonization setting | 0 to 100 percent on Precise Composite | No such parameter in the docs |
How do I keep the product unchanged?
Say so in the prompt and use a mask over the area that may change. The docs do not promise pixel-exact preservation, so inspect the result against your source. If you are choosing between models for packshots, this router post covers the options.
Which models accept a mask?
Only the ones that list it. The parameter table ties mask_url to ChatGPT Image 2.5 edits, and a parameter a model does not list is rejected with 400 unsupported_parameter. Check supported_parameters in the catalog first.
Sources
Related posts
More in Use cases
- FLUX Virtual Try-On v2: 4 MP inputs and a Sume reference edit
BFL's vto-v2 keeps inputs up to 4 MP as-is. Sume has no try-on endpoint in these docs; a garment swap is a reference edit with public HTTPS images.
- FTC: actors, dramatizations and scripted AI avatar ads
The FTC Q&A says actors in an obviously fictional dramatization are not giving testimonials, yet could still be deceptive. What that means for avatar ads.
- Is #ad enough in a video? What the FTC Q&A says
The FTC says #Ad may work at the start of a text post but may be too easy to miss in a video. Text in a video must stand out. Sume captions burn it in.
- FTC reviews rule and AI stock avatars: fake or false testimonials
The FTC Q&A says no blanket ban on AI avatars; a testimonial is barred only if fake or false. Here is how a Sume avatar and script map to that answer.
Written by Sume