At its annual MAX conference, Adobe announced new video capabilities to its Firefly AI platform, including a beta version of its Firefly Video Model which it previewed earlier this year. This marks Adobe's official entry into AI-powered video creation, joining its existing image and vector generation tools.
Here are the key capabilities of the new Firefly Video Model:
- Generative Extend: Available in the Premiere Pro beta, this tool allows editors to artificially lengthen video clips. It generates new frames to fill gaps, smooth transitions, or hold shots for longer durations.
- Text to Video: Users can generate video clips from text prompts, with options to adjust camera angles, motion, and zoom.
- Image to Video: This feature animates still images, turning them into short video clips.
These tools aim to address common challenges in video editing, particularly in adjusting timing and creating b-roll footage.
Adobe has also updated its existing Firefly models. The Image Model now generates content up to four times faster, while the Vector Model in Illustrator offers more control over pattern density in designs.
For enterprise users, the company announced AI-powered dubbing and lip-sync capabilities, as well as bulk editing tools for managing large volumes of content.
The Firefly Video Model is now available in limited public beta, with free access during this phase. The company hasn't released any details on pricing as yet.