New Firefly updates boost realism, control, and collaboration across Creative Cloud apps
Adobe just dropped major AI upgrades:
- Firefly Image Model 4 and Model 4 Ultra now power text-to-image generation
- A new collaborative tool, Firefly Boards, enters beta
- Third-party AI integrations join Adobe’s creative suite
- Creative Cloud apps like Photoshop and Illustrator get more AI magic
This is Adobe’s boldest move yet to solidify Firefly’s role in the generative AI race and to keep up with OpenAI and Google. The push adds speed, control, and high-res output to image generation, plus new tools for creative collaboration.
What’s new and powerful:
Firefly Image Model 4 is now Adobe’s fastest and most controllable model. It allows creators to generate 2K-resolution images with customizable styles, angles, and formats.
For more complex, hyper-detailed images, Model 4 Ultra promises better realism and structure rendering and thinks of intricate scenes or photorealistic elements.
Creative freedom grows:
The new Firefly web app now includes access to third-party models like OpenAI’s GPT image model, Google’s Imagen 3, and Veo 2 for video. More integrations are coming, including Luma, Pika, and Runway.
However, Adobe is clear: only its own models are marked “commercially safe.” Why? They’re trained on licensed and public content, unlike some rivals facing copyright scrutiny.
A new way to collaborate:
With Firefly Boards, Adobe is turning moodboarding into a shared generative AI experience which is similar to Figma’s FigJam but powered by AI. A mobile app is also on the way.
Photoshop and Illustrator aren’t left out:
- Photoshop adds smarter color tools and editing suggestions
- Illustrator brings generative shape fills and text-to-pattern features out of beta
Adobe isn’t just adding AI. It’s building a full AI ecosystem across devices and platforms and it wants to be the safest, most creator-friendly option out there.
Stay tuned: Adobe’s upcoming AI creative agent could reshape how pros use Photoshop.
