Adobe declares that its latest image generation model is the most advanced yet

Adobe’s Firefly family of generative AI models has been a topic of discussion among creatives, with the Firefly image generation model receiving criticism for its flaws and shortcomings compared to its competitors. However, Adobe is looking to change that narrative with the release of its third-generation model, Firefly Image 3, at the Max London conference this week.

The new model, now available in Photoshop (beta) and Adobe’s Firefly web app, promises more realistic imagery with improved capabilities in understanding complex prompts and scenes, enhanced lighting, and text generation. Adobe claims that Image 3 is better at rendering typography, iconography, and detailed features of people with various moods and expressions.

In a comparison between Image 3 and its predecessor, Image 2, the improvements are evident in terms of detail, lifelikeness, and overall expressiveness. The new model seems to offer more detailed and realistic outputs, showcasing advancements in shadowing, contrast, and scene understanding.

Apart from enhanced text-to-image features, Image 3 also powers new tools in Photoshop such as a style engine, auto-stylization toggle, and generative tools like Generate Background, Generate Similar, and Enhance Detail. These tools allow for precision edits, background replacement, and style transfer in images.

Despite the upgrades, Adobe has confirmed that the pricing for Firefly image generation will remain the same for now, along with its generative credit system and indemnity policy. The company continues to focus on providing a seamless and ethical AI-generated content creation experience for its users.