Adobe adds more image generators to its growing AI family

Adobe Unveils Two New Generative AI Models Alongside Exciting Updates

Adobe has just unveiled two new iterations of its groundbreaking text-to-image generative AI model, Firefly Image, along with a slew of other exciting features across its Creative Cloud apps. This launch follows the footsteps of industry leaders like OpenAI and Google, offering users a dual-track approach to creativity—speed and efficiency versus depth and detail.

Meet Firefly Image Model 4

According to Adobe, Firefly Image Model 4 is the fastest, most controllable, and most realistic version of the Firefly series yet. It empowers creators to produce images in up to 2K resolution, with enhanced control over factors such as style, format sizes, and camera angles. The improvements over its predecessor focus on delivering high-quality outputs swiftly and efficiently. For projects requiring greater detail and realism, Adobe has introduced Firefly Image Model 4 Ultra. This advanced model excels at rendering intricate scenes with small structural elements, ensuring every element shines with clarity.

Impressive Examples

Take a look at these captivating AI-generated images—a child, flowers, and a tiger—that showcase the capabilities of the new models. While the child’s face does have slightly odd shadowing, the overall quality remains stunning. Adobe hasn’t disclosed which specific models were used for these examples, but they demonstrate immense potential.

Image: Adobe

Firefly Web App Gets a Boost

The Firefly web app is now home to not only the new image models but also Adobe’s text-to-video and text-to-vector models, both previously released in public beta. A brand-new addition today is Firefly Boards, a collaborative generative AI moodboarding app reminiscent of FigJam. It was initially teased as “Project Concept” during Adobe’s Max event in October. Additionally, Adobe promises a Firefly mobile app for iOS and Android is on the horizon.

Opening Up Third-Party Options

For those eager to explore beyond Adobe’s ecosystem, the Firefly web app now supports third-party AI models. Users can choose from options like OpenAI’s GPT image model or Google’s Imagen 3 for images, and Google’s Veo 2 for video. Support for additional platforms like Luma, Pika, Runway, fal.ai, and Ideogram is slated to arrive soon. However, Adobe emphasizes these third-party models are intended for experimentation rather than professional use, distinguishing its own offerings as “commercially safe.” Given Adobe’s reliance on public or licensed content, this distinction makes sense compared to competitors who lack similar assurances.

Image: Adobe

Updates Across Creative Cloud Apps

Adobe isn’t stopping at AI innovations. Its popular Creative Cloud apps are getting refreshed too. Illustrator’s generative shape fill and text-to-pattern tools are now fully available following their beta debut last year. In Photoshop, users can expect streamlined color adjustments and automatic selections for hair, clothing, and facial features. The Actions panel is also receiving a makeover, offering tailored editing suggestions based on individual styles. This update lays the groundwork for Adobe’s upcoming creative AI agent, set to integrate seamlessly into Photoshop’s editing capabilities.