Adobe Firefly is taking a major leap from image generator to full creative studio, now adding multitrack video editing, AI-generated soundtracks and voiceovers directly inside its browser-based platform.
Unveiled at the AdobeMAX 2025 conference in Los Angeles, the update means you can create, edit and finish a whole project in the one place, without jumping into separate apps.
It also raises a big question for creators: how much of your workflow could now be handled in Firefly instead of Adobe’s more complex desktop tools?
Video editing now built directly into Firefly
One of the headline additions is a new private beta Firefly video editor, offering creators precision timeline editing in the browser. You can organise clips, generate assets on the fly, trim, sequence and add titles. It supports frame accurate control and even text based edits through transcript workflows.
Users can bring in their own footage and combine it with generated shots, including stylised looks like claymation or anime. Title templates and quick generation tools make scratch edits fast, while a history panel keeps track of previously created media.
The Firefly video editor could be a handy way to quickly edit videos on the go, as well as fill in missing shots or effects using generative AI, without launching into Premiere Pro. Speaking of Premiere, Adobe also announced its Premiere for Mobiles app, you can read more on this here.
Generate soundtracks and voiceovers in seconds
Firefly now includes AI audio creation tools designed for licensed, commercially safe output. Generate Soundtrack produces multiple music variations that sync to your video automatically. Generate Speech adds lifelike text to speech voices with fine control over emotion, pacing and emphasis.
Adobe has partnered with ElevenLabs for multilingual voice options, helping creators rapidly produce content for different audiences.
These new audio creation options can be especially useful for creating copyright safe music and voice overs, or simply spending less time and money using stock licences.
Firefly Image Model 5 boosts photorealism
Adobe’s latest model, Firefly Image Model 5, generates native 4MP visuals with improvements in fine detail, accurate lighting and realistic human anatomy. It also introduces conversational editing with Prompt to Edit, allowing users to describe changes to an image using natural language.
Interestingly, the Image Model 5 announcement was quite short, with a much greater spotlight on Adobe’s integration of many different 3rd-party generative AI models into its apps and platforms.
Personalised content with custom models
Firefly Custom Models let creators train AI using imagery they have rights to use, producing asset sets that stay consistent to their brand or style. Models are private by default and can only be shared with permission, making this a safer approach for commercial creative teams.
This was demonstrated at the keynote and it was surpassingly easy to train a custom model using existing assets, and looks to be a useful way for many businesses to add consistency to their artwork.
More AI models to choose from
Firefly continues to add partner models including Google, Topaz Labs, Black Forest Labs, OpenAI, Runway and more. This means creators can access a library of top AI engines within one subscription rather than jumping between services.
So now, you can simply select which model is particularly strong at a particular type of generative work, such as adding photo-realistic visual effects to a video.
It’s also an interesting admission by Adobe that its own generative AI models may not always be the right tool for the job all the time, allowing you a more seamless pathway to the best results.
Project Moonlight previewed across Adobe’s ecosystem
Another previewed feature, Project Moonlight, is an AI creative assistant that can analyse content and social performance to help with ideation and recommendations. Once finished, Adobe says creators will be able to move from concept to finished content in minutes simply by describing what they want.
Adding the ability to analyse the performance of your social posts is particularly interesting, and may help improve views and engagement of your future posts.
Availability and pricing
Many features such as Prompt to Edit, Generate Soundtrack and Generate Speech are available in public beta now. Others like the Firefly video editor and Custom Models are in private beta, with waitlists open.
Creative Cloud Pro and Firefly plan customers currently have access to unlimited image and video generations through to 1 December.
What this all comes down to is that you’ll be able to do more in Adobe Firefly, as it’s not just the place to go for creating generative AI images. It could also mean avoiding, or even opening Adobe’s more pro-level apps like Premiere Pro, Audition, Photoshop and more.
The post Adobe Firefly becomes your all-in-one AI tool for creating anything appeared first on GadgetGuy.
0 comments:
Post a Comment