Adobe Firefly capabilities
5 mapped capabilities, each graded and dated. The map shows what Adobe Firefly can do; the audit shows whether it’s worth consolidating — and a guide shows how to move.
Capabilities
Commercial safety and Content Credentials
canonicalverified todayAdobe's own Firefly models are positioned as 'designed to be commercially safe': trained on Adobe Stock licensed content, openly licensed content, and public-domain material, with no customer content used for training. Generated assets carry Content Credentials provenance metadata, and qualifying plans add IP indemnification for Firefly-generated outputs.
Generative Fill and Generative Expand
canonicalverified todayFirefly edits existing images by adding, replacing, or removing objects within a masked region (Generative Fill) and by extending the canvas beyond its original borders with newly generated, context-aware content (Generative Expand).
Plans and generative credits
canonicalverified todayFirefly is sold as standalone subscription tiers metered by a monthly 'generative credit' allowance, in addition to a free tier and bundling inside Creative Cloud plans. Standard generations on Adobe's own models are unlimited on paid plans; credits are consumed by premium features such as video, translation, and partner models.
Text-to-image generation
canonicalverified todayAdobe Firefly turns a natural-language text prompt into images from the Firefly web app (firefly.adobe.com), with selectable models, aspect ratios, content/structure reference images, and style controls. Adobe's own flagship is the Firefly Image Model 5, which generates at native 4MP resolution and targets photorealistic detail and anatomically accurate portraits.
Third-party (partner) model access
canonicalverified todayBeyond Adobe's own Firefly models, the Firefly app exposes a curated, growing roster of third-party image and video models so creators can pick the best engine per project from a single interface. The lineup has included Google Gemini (Nano Banana), Black Forest Labs FLUX, Luma AI Ray, Runway, OpenAI GPT Image, and others, with 30+ models across the platform.