Firefly Services operation definition overview

Last updated on Feb 20, 2025

Applies to enterprise.

This document describes the available APIs and capabilities for Firefly, Photoshop, Lightroom, Video & Audio, and Content Tagging (powered by Sensei Platform) as part of the Firefly Services offer.

Firefly APIs

API action	Capability	Operation definition (1)
Text to Image	Generate images based on a text prompt	1 Operation = 1 standard image generated with foundation Firefly model
Generative Fill	Fill the masked area of an image with an optional prompt	1 Operation = 1 standard image generated with foundation Firefly model
Generative Expand	Expand an image to a new size/aspect ratio	1 Operation = 1 standard image generated with foundation Firefly model
Text to Image with Style Reference	Generate images based on a prompt with optional reference image to match style	1 Operation = 1 standard image generated with foundation Firefly model
Text to Image with Structure Reference	Generate images based on a prompt with optional reference image to match structure.	1 Operation = 1 standard image generated with foundation Firefly model
Generate Similar Images	Generate similar Images based on the reference image.	1 Operation = 1 standard image generated with foundation Firefly model
Generate Object Composite	Upload an image (with or without mask), such as a product photo, and utilize a text prompt to generate a seamlessly composited scene featuring the product.	1 Operation = 1 standard image generated with foundation Firefly model

(1) A standard image means an image width of up to 2000 x 2000 pixels. API requests can result in multiple images being generated. Pricing does not cover any future capabilities to generate imagery using “custom model extensions”.

File Size Limits

For image assets used for Generative Fill, Generative Expand, or Generative Match, the File Size Limit per successful API call is up to 15 megabytes (or MB).

Creative Cloud APIs

Photoshop APIs

API action	Capability	Operation definition (2)
Remove Background	Isolate subject of interest in an image and remove background	1 Operation = 1 API call per asset
Create Mask	Isolate a subject of interest in an image (people, objects, etc) and generate an image mask	1 Operation = 1 API call per asset
Create Renditions	Create flat image representations of a PSD	1 Operation = 1 API call per asset
Create actionJSON	Convert .atn file to an actionJSON format	1 Operation = 1 API call per asset
Execute actionJSON	Execute a Photoshop Action with actionJSON format	1 Operation = 1 API call per asset
Execute Photoshop Actions	Executes Photoshop Action file against a PSD, JPEG, PNG, or TIFF	1 Operation = 1 API call per asset
Replace Smart Object	Replace Smart Object in a PSD	1 Operation = 1 API call per asset
Product Crop	Smart crop an image keeping the subject of interest in view	1 Operation = 1 API call per asset
Depth Blur	Apply depth blur to an image input	1 Operation = 1 API call per asset
Edit Text Layer	Change the contents of a text layer in a PSD, e.g. for localization	1 Operation = 1 API call per asset
Apply PSD Edits	Apply basic layer edits (name, state, etc), add/edit adjustment, pixel, and shape layers	1 Operation = 1 API call per asset
Create New PSD	Create new PSD with layers	1 Operation = 1 API call per asset
Create an Artboard	Create artboards from multiple PSD inputs	1 Operation = 1 API call per asset
Get Manifest	Extract metadata from PSD document, including general file and layer information	1 Operation = 1 API call per asset

Lightroom APIs

API action	Capability	Operation definition (2)
Straighten Image	Auto-straighten an image	1 Operation = 1 API call per asset
Autotone	Apply autotone to an image	1 Operation = 1 API call per asset
Edit Image	Adjust Lr adjustments to a photo, including exposure, contrast, detail, etc.	1 Operation = 1 API call per asset
Apply Preset	Apply Lr preset to a photo	1 Operation = 1 API call per asset
Apply XMP	Add XMP metadata to a photo	1 Operation = 1 API call per asset

Content tagging APIs

API action	Capability	Operation definition (2)
Color Tagging	Given an image, can compute a histogram of pixel colors and sort them by dominant colors into buckets	1 Operation = 1 API call per asset
Keyword Tagging	Given a text document, the keyword tagging service automatically extracts keywords or key phrases that best describe the subject of the document	1 Operation = 1 API call per asset
OCR	Given an image, can indicate if text is present in the image. If text is present, OCR can return the text	1 Operation = 1 API call per asset

(2) API Actions can be run for multiple assets. As an example, if a Customer requests an API Action for one hundred (100) assets, then such API Action would be counted as one hundred (100) Operations. For clarity, status polling API actions don't consume Operations.

File size limits

For image assets (For example, PSD and PNG files), the Creative Cloud API File Size Limit per successful API call is up to 300 megabytes (or MB).

In the Customer’s use of the Photoshop and Lightroom APIs, the file size per successful API call for an asset is limited to the file size limit per successful API call as set forth for each asset type. If the Customer exceeds the file size limit per successful API call, the Customer will be charged for an additional successful API call.

For example, a 200 MB PSD file would count as one successful API call, a 500 MB PSD file as two successful API calls, and a 700 MB PSD file as three successful API calls.

Additional

Firefly Creative Production

API action	Capability	Operation definition
Remove Background (flat image output)	Isolate the subject of interest in an image, remove the background, and output a flat image file	1 Operation = 1 output file
Remove Background (layered PSD output)	Isolate the subject of interest in an image, remove the background, and output a layered PSD file	3 Operations = 1 output file
Crop	Crop image sizes or aspect ratios	2 Operations = 1 output file

Get help faster and easier

New user?