The future is Firefly: Unlock new levels of creativity with the latest generative AI innovations

Image generated using Adobe Firefly.

Adobe Firefly empowers anyone — at any design skill level — to dream, have fun, and express themselves by creating anything they can imagine with simple text prompts. Users have generated over 3 billion images with Firefly’s initial model since its beta launch in March, making it the most popular AI image generation model designed for safe commercial use, in record time, globally.

Firefly is built with the grassroots support and participation of our community, and along the way, we’ve listened closely to your feedback on how best to shape the future of generative AI for creativity. We’re proud of what we’ve achieved thus far, and excited for the new innovation coming your way!

Firefly’s amazing journey so far

Firefly has come a long way already, let’s take a look at the main milestones we’ve crossed this year:

• Firefly is now available in Photoshop (try Generative Fill and Generative Expand, also in your browser with Photoshop on the web), Illustrator (try Generative Recolor), Adobe Express (try Text to Image and Text Effects) and as a standalone web app at firefly.adobe.com.
• Outputs from most Firefly modules and features can now be used commercially. This includes any type of usage that generates commercial activity for you, your brand, or your business.
• Firefly now supports text prompts in over 100 global languages, while the Firefly web app is available in 20 languages, with more options being added in the coming months.

And we’re continuing to evolve: new features, more creator control

At Adobe MAX, we’ve made a number of major announcements that will take Firefly users a step further by making it easier and faster to realise their creative vision. Here’s a quick overview:

Introducing the Adobe Firefly Image 2 Model

The new Adobe Firefly Image 2 Model, now available as beta on the Firefly web app, significantly advances creator control and image quality. The next chapter of generative AI for imaging provides larger model capacity and better image generation — making it faster and easier to realise your creative vision and tell your story with high quality outputs. And you can easily toggle between the Image 1 and Image 2 models to see the difference with your own eyes.

Text to Image enhancements

Firefly Image 2 powers a range of new Text to Image features in beta that you can now try out in the Firefly web application on desktop or mobile to more easily experiment and explore ideas:

• Generative Match: Apply a style from a pre-selected set of images or upload your own style reference image to easily create new ones that represent your style or brand. Create consistent content at scale by recreating the look and feel of one of your images across different assets. Combine these image styles with the included Style Effects like “golden hour” or “photographed through window” for a great effect.   Read more here.
• Photo Settings: Apply and adjust photo settings, similar to manual camera controls, to achieve a more photorealistic image quality with higher-fidelity details (including skin pores and foliage) and effects like motion blur and depth of field.
• Prompt Suggestions: Learn how to expand or reword prompts to generate superior results faster – we do the heavy lifting and propose prompts designed to generate beautiful outputs.
• Share from Firefly and Save to Libraries: Share links to generated images and their prompts directly from Firefly to help others understand how they were created and what prompts were used to make them. With shared links others can also remix and iterate on your work, making collaboration easier. Also, save Firefly generations to Creative Cloud Libraries so you can reopen them within all the Creative Cloud apps for collaborative workflows.

Introducing the Adobe Firefly Vector Model in Adobe Illustrator

As of today, you’re able to use Text to Vector Graphic as a beta feature in Illustrator on the desktop to easily create a wide array of fully scalable and editable high-quality vectors — including icons, scenes, subjects, and patterns — starting from a simple text prompt.

This groundbreaking new feature, powered by the new Firefly Vector Model (the world’s first generative AI model for vector graphics), generates vectors designed to be edited by a designer – few points, gradients, grouped layers and amazing editable gradients. It brings together our generative AI expertise and our decades of vector innovation to help you jumpstart your next project or enhance your work with new graphics that perfectly match the style of your existing design.

Introducing the Adobe Firefly Design Model in Adobe Express

From today you can also try out Text to Template as a beta feature in Adobe Express, a feature powered by the new Firefly Design Model (which was trained on hundreds of thousands of high-quality Adobe Express templates) that makes getting started with a design even easier. Instantly generate stunning templates from a simple text prompt to create social posts, flyers, posters, and more within minutes.

Templates generated with Text to Template are fully editable in Adobe Express and can be custom designed in all popular aspect ratios for print, social, and web. Text is automatically placed and sized to help send core messages from prompts to audiences, and text colors are intelligently selected to maximise readability.

More Firefly news

Generative Fill in Adobe Express

In addition, users can now also use Generative Fill — which debuted in Photoshop earlier this year — in Adobe Express, to easily insert, remove, or replace objects or people in any design with a simple text prompt.

Setting the standard in creative generative AI

Firefly was built specifically with creators in mind. It empowers users of all experience levels, across our apps, to enhance their workflows and realise their creative vision — faster and easier.

Creator choice and control are therefore paramount, along with helping to ensure outputs are safe for commercial use, so you can design with confidence. To help promote transparency and trust, Content Credentials (a new kind of tamper-evident metadata available through the Content Authenticity Initiative) are by default automatically included on AI-generated content made or modified using Firefly.

The Adobe Stock contributor community of photographers, designers, illustrators, filmmakers, and multimedia artists of all kinds is vital to the evolution of generative AI, and to demonstrate that, we have developed a compensation model — the Firefly Contributor Bonus — that ensures creators can benefit from this revolutionary new technology and enjoy practical advantages.

Finally, Adobe is committed to ensuring the safety and responsible development of our AI technology. We are constantly and rigorously testing our models to ensure they don’t perpetuate harm and bias. Our AI features also have feedback mechanisms built in so users can report any concerns. Based on this input, our engineering team then resolves any issues to make generative AI better for everyone.

Go ahead, create something extraordinary with Firefly

Try all the new Firefly features and experiment with them in your workflow by updating your apps to the latest version or play with them. It’s now easier than ever to create outstanding content with the power of generative AI. What will you create next?

We want to hear from you, so join the conversation on Discord or the Adobe Firefly Community forum to exchange ideas, share your creations, and provide feedback.