Adobe Firefly now supports prompt-based video editing, adds more third-party models

Adobe has unveiled exciting updates to its AI-driven video creation tool, Firefly, introducing a sophisticated video editor that enhances the editing experience with prompt-based capabilities. This new functionality allows users to make specific edits to video elements, such as colors and camera angles, without the need to start from scratch if a particular aspect doesn’t meet their expectations. Previously, Firefly only facilitated prompt-based generation, requiring users to regenerate entire clips for minor alterations. The newly introduced timeline view simplifies the editing process, enabling users to adjust frames and sounds seamlessly. Initially revealed in October as part of a private beta, this feature is now accessible to all users. With the integration of Runway’s Aleph model, users can provide detailed instructions to Firefly, enabling commands like "Change the sky to overcast and lower the contrast" or "Zoom in slightly on the main subject." Additionally, Adobe’s own Firefly Video model allows for innovative features like uploading a reference video to replicate camera movements effectively. Further enhancing its capabilities, Firefly now supports the Topaz Labs’ Astra model, which enables the upscaling of videos to resolutions of 1080p or 4K. The addition of Black Forest Labs’ FLUX.2 model for image generation, alongside a new collaborative boards feature, broadens the tool’s appeal. The FLUX.2 model will be immediately available on Firefly across all platforms, with Adobe Express users expected to gain access in January. As competitors continue to innovate in the realm of image and video generation, Adobe aims to attract more users to its platform. To incentivize engagement, subscribers of Firefly Pro, Firefly Premium, and those with credit plans will enjoy unlimited generations from all image models, including the Firefly Video Model, until January 15. This year, Adobe has significantly enhanced its Firefly offerings, launching a subscription model for varied levels of image and video generation, a new web app, and mobile applications, while continuously integrating third-party models to enrich the user experience.