Here’s one way to use it. You can upload a photo of yourself, along with a recording of your voice. Then, you type a prompt about what you want the digital version of you to say and do. The system then bases the video on the photo, the voice on your sample, and the action on your instructions.
The model can create full-body movement, not just faces, and accurately matches lip and facial motion to the audio. It handles up to 90 seconds of video in one go, with output up to 4K at 60 frames per second.
Character-3 is used by content creators, marketers, teachers, and businesses. For example, teachers can create videos with virtual instructors, and brands can utilize digital spokespersons for their advertisements. The tool is part of Hedra Studio, a web platform that brings together Character-3 and other AI tools for video, audio, and images. Hedra Studio enables users to build scenes, add backgrounds, and control the look and feel of each video. Developers can also use Character-3 through an API.