Adobe’s Firefly AI is now available as mobile apps for iPhones and Androids, the company announced on Tuesday. These apps are free to download and let you use Firefly to create AI images and videos on the go. Plus, the app comes with a few free generative credits for you to experiment with Adobe’s AI.
Adobe is also expanding its roster of third-party AI partners to include six new models from Ideogram, Pika, Luma and Runway. Google’s latest AI models are also joining the lineup, including the internet-famous Veo 3 AI video generation model with native audio capabilities and the Imagen 4 text-to-image model support. Finally, its moodboarding AI program, Firefly Boards, is generally available today after months in beta.
Here’s everything you need to know about Adobe’s newest batch of Firefly AI updates. For more, check out our favorite AI image generators and what to know about AI video models.
Firefly AI for iOS and Android users
Adobe’s Firefly mobile apps will let you access its AI image and video capabilities from your phone. A mobile app felt like the next natural step, since Adobe saw that mobile web usage of Firefly noticeably increased after Adobe’s Firefly video capability launched in early 2025.
Not every Firefly feature will be available at launch, but for now, we know that these features will be included: text-to-image, text- and image-to-video, generative fill, and generative expand. You can download the app now from the Apple App Store and Google Play Store.
The app is free to download, but you’ll need a Firefly-inclusive Adobe plan to really use the app. In the hopes that you’ll sign up for a full plan, Adobe gives you 12 free generation Firefly credits (10 for images, two for videos, which doesn’t shake out to many of each). So you can use those to see if Firefly is a good fit for you.
Firefly plans start at $10 per month for 2,000 credits (about 20 videos), increasing in price and generation credits from there. Depending on your Adobe plan, you may already have access to Firefly credits, so double-check that first.
Adobe’s six new AI models from Google, Runway and more
Adobe’s also adding new outside AI creative models to its offerings, including image and video models from Ideogram, Pika, Luma and Runway. You might recognize the name Runway from its deal with Lionsgate to create models for the entertainment giant. Ideogram, Pika and Luma are all other well-known AI creative services. Google’s Veo 3 AI video generator is also joining, bringing its first-of-its-kind synchronized AI audio capabilities, along with the latest generation of Google’s AI image model.
This is the second batch of third-party models that Adobe has added to its platform. Earlier this spring, Adobe partnered with OpenAI, Google and Black Forest (creator of Flux) to bring the companies’ AI models to Adobe. What’s unique about this is that all third-party models have to agree to Adobe’s AI policy, which prevents all the companies from training on customers’ content — even if the individual companies don’t have that policy on their own, it’s standardized across all models offered through Adobe. This is also true for the new models added today. For AI-wary professional creators who make up the majority of Adobe users, that’s a bit of good news.
You’ll need a paid Firefly plan to access outside models; otherwise, you’ll just have access to the Adobe models. Here are all the AI models available through Adobe:
Adobe Firefly Image 3Adobe Firefly Image 4Adobe Firefly Image UltraFlux 1.1 ProFlux 1 KontextGoogle’s Imagen 3OpenAI’s image generation model(new) Ideogram 3(new) Google’s Imagen 4(new) Runway’s Gen-4 Image
For video, you can use:
Adobe Firefly VideoGoogle Veo 2(new) Google Veo 3(new) Luma AI Ray 2(new) Pika’s text-to-video generator
Adobe’s own Firefly AI models are trained on a combination of Adobe Stock and other licensed content. You can learn more in Adobe’s AI guidelines and approach to AI.
AI moodboarding gets a boost
Other Adobe updates include the general release of its moodboarding program, Firefly Boards, which has been in beta since April. Moodboarding is a practice that lets you cluster together different elements, like colors and shapes, to evoke specific moods and aesthetics. It’s a good initial step for planning content and campaigns.
You can use the infinite canvas to brainstorm and plan content. You can generate images and videos in Boards using Adobe and non-Adobe models; the setups are very similar to generating in the regular Firefly window. Boards are collaborative, so you can edit with multiple people. A new one-click arrange button can help you organize and visualize your files more easily, a much-requested feature that came out of the beta.
Firefly boards are synced up with your Adobe account. So you can select a photo in a Board, open it in Photoshop and edit it. Those changes will then be synced up with your Firefly Board in less than a minute, so you can always see the latest version of your file without needing to be limited to editing in Boards.
For more, check out Premiere Pro’s first generative AI feature and the best Photoshop AI tools.