Adobe announced three new generative AI models that add powerful features to Illustrator and Adobe Express and greatly improve Photoshop’s text-to-image capabilities. At the Adobe Max event, Adobe introduced Firefly Image 2, the latest version of the original Firefly AI image creator that powers popular Photoshop features like Generative Fill, as well as two new Firefly models for creating vector images and design templates.

Firefly Image 2

Adobe says the new Firefly Image 2 model produces significantly higher quality images than its predecessor in areas such as foliage, skin texture, hair, hands and facial features, especially when creating photorealistic people. Images produced using the Firefly Image 2 model have higher resolution, more vibrant colors and color contrast.

The Image 2 model also comes with new AI-powered editing capabilities to help users customize their results. Photo settings can be applied to manually or automatically adjust the depth of field, motion blur, and field of view of a rendered image, just like manual camera controls. A “Prompt Routing” feature has also been added to automatically complete prompts to improve the efficiency of entered prompts.

Generative Match

Adobe also introduced a new “Generative Match” feature to match the style of rendered content with specific images. Users can copy the style by choosing from a pre-selected list of images or uploading their own references. Additionally, the strength of the similarity to the reference style can also be adjusted.

For the Generative Match feature, Adobe is also developing new policies and security measures to prevent abuse. The feature will ask users to agree to Adobe’s terms of use and confirm that they have the rights to use the uploaded image, as well as providing a level of accountability by storing a small snapshot of the uploaded content (not used to train AI models) on Adobe’s servers. The Generative Match feature will remain in beta until it is fully developed and will not be available for commercial purposes during this period. However, as it stands, it doesn’t seem like much is being done to prevent an artist’s content from being imitated. At the moment this system appears to be Adobe’s self-protection rather than preventing copycat behavior in the first place.

Firefly Image 2, on the other hand, can be tried today via the web-based Firefly beta and is coming to Creative Cloud apps “very soon.” So it won’t be available in Photoshop (standard, beta, or web) yet.

Adobe Illustrator and Firefly Vector

Adobe also announced the new Firefly Vector tool for Adobe Illustrator, which it claims is “the world’s first generative AI model for vector graphics.” The Firefly Vector model, currently available in Firefly beta, allows users to create editable vector images using text prompts that automatically separate each element of the graphic into “logical” groups and layers. Unlike traditional JPEG and PNG files, vector graphics (also known as SVG files) are used by logo designers because they can be scaled to any size without affecting the overall image quality.

Adobe says that, like the original Firefly text-to-image model, the Firefly Vector model is designed to be safe for commercial use (when it comes out of beta). The Firefly Vector model is available to try today via the Adobe Illustrator beta version.

Firefly Design

Adobe recently introduced Firefly Design, which creates customizable templates for print, social posts, online advertising, video and more. Powering the new text-to-template beta feature in Adobe Express, the Firefly Design model uses text prompts to create fully editable templates for “all popular aspect ratios.”

It bears some similarities to Canva’s Magic Design feature (another all-in-one design platform that rivals Adobe Express), in that users can create unique templates, such as “Hand holding a DH logo,” rather than dropping individual text and image elements onto a blank canvas. It can define something.

Although these new models are only available in beta for now, judging by the release timeline of Photoshop’s Generative Fill feature, it won’t be long before it becomes generally available. The original Firefly model has been used to create more than 3 billion images to date, according to Adobe, and it’s unusual for such a popular product to be updated before its first birthday.