Adobe Firefly AI can now create vector graphics in Illustrator

Adobe Illustrator, a vector graphics editing and designing tool for graphics artists, is about to enter the generative AI era with the release of the Firefly Vector Model.

Adobe describes the new model as "the world's first generative AI model focused on producing vector graphics." Similar to Firefly for creating images and photos, Firefly for Illustrator will be capable of creating entire vector graphics from scratch. Like other Firefly models, the vector model was also trained on data from Adobe Stock.

Illustrator's beta version now allows users to create entire scenes using a text prompt. The interesting thing is that these scenes can contain multiple objects. This is not just a jumble of vectors that make up the overall graphic; Illustrator will automatically generate these different objects, and users can manipulate them individually, like any other group or layer in Illustrator.

Adobe's VP for generative AI and Sensei, Alexandru Costin, told Techcrunch that the company used tens of millions of vector images in Adobe Stock to train Firefly, a new AI model, to create vector drawings.

Costin described the process as "a journey" because there has been less research on using generative AI to create vector drawings than on other types of images. This required more work from the team, who focused on creating a model that could generate these images with the fewest possible points.

A new feature called Mockup is coming to Illustrator, which will allow users to apply vector art to any 3D scene. This could be used to design a drink can or a mockup of a new logo on a t-shirt. Costin explained that Mockup is a great way to show customers what they are buying when they contract a freelancer, as it allows them to see the art in context.

A new tool, Retype, is now available. It converts static text in images to editable text and can also find matching fonts.

Additionally, Illustrator is now available on the web.