OmniHuman generates video using an image of a human figure paired with an audio file.
No supported capabilities recorded.
https://fal.run/video
/v1/models/fal/fal-ai/bytedance/omnihuman