Firefly Image 3 powers several new features in Photoshop, including a new style engine and an auto-stylization toggle, which allow the model to generate a wider array of colors, backgrounds, and subject poses. Other new tools include Generate Background, Generate Similar, and Enhance Detail, which leverage Image 3 to perform precision edits on images. The Firefly web app is also getting new features, Structure Reference and Style Reference, to coincide with the release of Image 3. Despite these upgrades, Adobe confirmed that its current pricing tiers and generative credit system will remain in place for now.
Key takeaways:
- Adobe is releasing its third-generation model, Firefly Image 3, which is designed to produce more realistic imagery than its predecessors, Image 2 and Image 1. The new model is available in Photoshop (beta) and Adobe’s Firefly web app.
- Firefly Image 3 has improved capabilities in understanding longer, more complex prompts and scenes, improved lighting and text generation, and more accurate rendering of typography, iconography, raster images, and line art.
- Adobe trains Firefly models on uploads to Adobe Stock, Adobe’s royalty-free media library, along with licensed and public domain content. The company also controversially trains Firefly models on AI-generated images, a practice some consider a form of data laundering.
- Firefly Image 3 powers several new features in Photoshop, including a new “style engine” and auto-stylization toggle, and three new generative tools: Generate Background, Generate Similar, and Enhance Detail. The Firefly web app is also getting new features, Structure Reference and Style Reference.