Adobe is leveling up its Firefly generative AI platform with a fresh set of tools designed to give creators more control, creative options, and flexibility, all from within the Firefly app. Rolling out now, the update brings beta support for text- and voice-generated sound effects, expanded video editing features, and access to a growing lineup of partner AI models, including Google Veo 3 and Moonvalley Marey.
Generate Custom Sound Effects With Text or Voice
Headlining the update is Generate Sound Effects (beta), a new feature inside the Firefly app that lets users create custom audio by simply typing a description or speaking it aloud. Whether it’s a cinematic thud, ambient rainfall, or the subtle hum of traffic, the tool supports impact sounds and atmospheric layers for more immersive content.
Firefly users can blend generated effects with uploaded footage, layering in ambience to set mood, location, or tone, without having to hunt through third-party libraries or fiddle with external tools. And yes, the feature works whether you’re chasing epic explosions or the delicate crunch of gravel underfoot.
More AI Models, One Seamless Workflow
Adobe is further deepening its Firefly partner ecosystem with support for several new third-party models, including:
- Moonvalley’s Marey
- Google’s Veo 3, now with audio generation
- Runway’s Gen-4 Video
- Pika 2.2
- Luma AI’s Ray2
- Topaz Labs’ image and video upscalers
All of these join Adobe’s existing in-app models from OpenAI, Black Forest Labs, Ideogram, and others. The appeal: creators can access everything from stylized visuals to high-res video upscaling without bouncing between multiple tools or accounts. All partner content respects Adobe’s content policies, user uploads and generated assets are not used to train models, and Content Credentials are applied automatically to indicate the model used.
One login, one plan, zero app-hopping.
Firefly Video Gets Precision Tools for Creators
The Firefly app now offers deeper control over video generation and editing with new frame-specific tools aimed at professionals working on everything from vertical social clips to long-form cinematic cuts. New capabilities include:
- Composition Reference: guide framing based on an existing image
- Keyframe Cropping: adjust layout across multiple frames
- Style Presets: quickly apply a consistent look across outputs
These tools are designed to save time while still allowing for detailed creative adjustments. This will be useful whether you’re building brand content, pitching client visuals, or simply fine-tuning a personal project.
One Platform, Billions of Creations
According to Adobe, Firefly users have generated over 26 billion assets to date using its family of AI models. With this latest update, that number is likely to accelerate as more creators explore what’s possible when sound, video, and visual design come together under one unified, customizable toolkit.
More details on the new features and model integrations can be found on the official Firefly site.
