At the regional Adobe Max London this week, Adobe got the crowd buzzing with a bevy of Firefly-centric announcements, including the new all-in-one Firefly app, new Firefly models including Firefly Image Model 4 and Image Model 4 Ultra, new Firefly features, a new Firefly mobile app, Firefly Boards, and more.
It seems like it was only yesterday (actually, it was just a few weeks ago) when Adobe opened the floodgates to its Firefly Video Model in general release, bringing innovative AI-powered features to all Premiere Pro users. At its Max event in London this week, Adobe showed that it has more Firefly innovations on tap, including the all-new, all-in-one AI Firefly app, serving as a centralized hub for AI-powered content ideation, creation, and production.
Calling it the next generation of creative AI, Adobe rolled out the new Firefly, which unifies AI-powered tools for image, video, audio, and vector generation into a single, cohesive platform. So now, creators can generate this range of content from a single location, iterate on their work across Adobe’s creative apps, and seamlessly deliver them into production—all within the new Firefly app.
The new Firefly also features enhanced models, improved ideation capabilities, expanded creative options, and unprecedented control.
Since its inception, the Firefly app has operated within the Adobe universe only, but that is changing, as Adobe is integrating non-Adobe models into the Firefly ecosystem, giving users a choice when it comes to integrating third-party partner models into their work. Users can experiment with models built by partners with different aesthetic styles in the Firefly app, starting with Google Imagen 3 and Veo2, OpenAI GPT image generation, and Black Forest Labs Flux 1.1 Pro.
Adobe says it will be easy to switch between models, with transparency provided as to which model is being used at each step in the process. The default option will be for Firefly models, and users will have to opt in to the non-Adobe models. Also, content credentials will automatically attach to all AI-generated content, clearly indicating whether it was created with an Adobe Firefly model or partner model.
New models, new features
Adobe makes a habit out of showing how popular its Firefly is, proudly displaying the latest numbers. In case you were wondering, the most recent stat from Adobe puts the total number of assets generated by Firefly at 20 billion-plus—not bad for being on the market less than two years. And no doubt, that number will continue to grow, given the work Adobe continues to do relating to its generative AI technology.
At this point, Adobe has a comprehensive creative AI platform, which it has been building model by model, feature by feature. This includes Image Model, Vector Model, Design Model, Video Model, Audio Model, and the soon-to-come 3D Model, powered by Firefly.
Earlier this year, Adobe unveiled the Firefly Web app and introduced new capabilities across video and audio. Now, Adobe has announced the latest commercially safe, production-ready Firefly models—Firefly Image Model 4 and Firefly Image Model 4 Ultra—upping the ante in terms of photorealism with lifelike detail, enhanced structure control, and 2K resolution. Adobe called this fourth installment the fastest, most controllable, and most realistic Firefly Image Model yet.
Back in the spring of 2023, Adobe introduced Firefly Image Model, following closely afterward with Firefly Image Model 2 and 3. Now, photorealism has taken a major step forward with Image Model 4, the next foundational update. Image Model 4’s features raise the bar for professional image generation. Firefly Image Model 4 offers lifelike image quality, more creative control over the structure and style of outputs, and more control over camera angles and zooms, as well as the speed to rapidly explore and iterate concepts.
With the ability to generate outputs and up to 2K resolution, creative professionals can crop and recompose images with greater precision and print in larger formats without sacrificing quality. According to Adobe, these new models deliver significant improvements in text rendering, as well as humans, animals, and architecture compared to results from Image Model 3.
And if this increased image quality is not enough for pro users needing more detail and realism for their projects, there is Image Model 4 Ultra.
Both the Image Model 4 and Image Model 4 Ultra are generally available now.
As noted earlier, the Firefly Video Model was set free into general availability earlier in the month and is available through Firefly on the Web. This is the industry’s first IP-friendly, production-ready video model, with tools creatives can use to generate video clips from a text prompt or an image; they can also use camera angles or specify start and end frames to control shots, craft atmospheric elements, develop custom motion design elements, and more.
Meanwhile, the Adobe Vector Model, generally available, powers the new text vector capability in Firefly, which is also now generally available, giving designers the ability to create high-quality, fully editable vector-based artwork, logo variations, product packaging, icon scenes, patterns, and so forth with simple text prompts written in everyday language.
As Adobe pointed out, the modalities, mediums, and models on Firefly continue to get better and better. In addition to the new Text to Vector, Adobe also is offering Scene to Image (beta), an update to the feature that launched last year for enabling a 3D shape to influence a user’s generation. It leverages some capabilities of Project Neo.
Firefly mobile app
In addition to the new Firefly Web app, Adobe also introduced the new Firefly mobile app, which it said is coming soon on both iOS and Android. The Firefly mobile app is for pros as well as new creatives, giving them the power of Firefly AI in their mobile device. The app can be used for just about any creative endeavor, from fast concepts to advanced generative AI creations, enabling users to refine colors, adjust compositions, or completely transform a design into something new.
It supports features such as Text to Image, Generative Fill, and Generate Video, and will automatically sync with the person’s Creative Cloud account. Projects created on the phone with Firefly mobile can be continued on the web or desktop.
Firefly Boards
Flying into public beta starting now is the new Firefly Boards, a multi-person infinite canvas app for organizing and visualizing all the content of a project within one place. Previewed last year as Project Concept, Firefly Boards provides a fast, easy way to brainstorm and explore and align ideas before heading into production, giving creators an AI-first surface for moodboarding, storyboarding, brainstorming, and exploring creative concepts. It is compatible with a user’s creative tools and is built for generative workflows, though it also works with non-generative tools as well.
The initial feature set includes:
• Multi-person collaboration
• Generation with Firefly 4, Google Imagen, Flux Pro 1.1 Ultra
• Generative ideation with Remix, Sample
• Generative editing with GenFill/GenExpand
• Robust layout and organization tools, including Adobe Fonts
• Send-to-Photoshop/Express
At Max London, Adobe also announced improvements to its Creative Cloud applications and more.
LIKE WHAT YOU’RE READING? INTRODUCE US TO YOUR FRIENDS AND COLLEAGUES.