Adobe has announced the beta release of its Firefly Video Model at the Adobe MAX event on October 14, 2024. This marks the expansion of Adobe's generative AI capabilities, previously focused on images and design.
The Firefly Video Model allows users to create videos using text-based commands, controlling camera angles, movements, and styles. This feature enables content creators to generate videos from descriptions, similar to how Firefly generates images.
A key feature is Generative Extend, now available in Adobe Premiere Pro, which allows users to extend video clip durations, smooth transitions, and fill gaps in footage without re-recording.
According to Adobe, Firefly is trained on fully licensed data, ensuring that generated content can be used freely by creators without copyright concerns.
Alongside the video model, Adobe introduced updates to existing image, vector, and design models within the Firefly ecosystem, including a fourfold increase in image generation speed.
Adobe Illustrator has also been enhanced with features like Generative Shape Fill and Generative Recolor, enabling designers to create unique patterns and add detailed elements quickly.
Creators interested in the Firefly Video Model can sign up for limited beta access, with free usage during this phase. Adobe will provide pricing details when the tool is fully available.