This model works without a text prompt

Select or upload an image with an avatar and a voiceover to generate a lip-sync.