lipsync

To synchronize facial mouth movements with audio for video dubbing, you provide an existing video and an audio track. The Skill routes the input through RunComfy models to generate a natural lip-synced video while preserving the subject's appearance and lighting.

Yes, you can generate a talking avatar from a portrait and a voiceover script. The Skill supports script-to-video generation, creating synchronized speech and matching facial animation even when no pre-recorded audio file is provided.

The best way to create multilingual voiceovers for existing footage is to pair the original video with translated audio tracks. The Skill applies audio-driven lip-sync to match the new speech with the subject's mouth movements, preserving the original background and motion.

Yes, this lip-sync tool works with stylized character animation. It synchronizes audio-driven facial animations for various formats, including portrait-based talking avatars and stylized characters, by routing inputs to compatible models like ByteDance OmniHuman or Kling.

For audio-driven video generation, you need accessible media URLs for your video or portrait, valid CLI authentication, and compatible audio and video durations. You must also obtain consent from the people whose faces or voices are used.

Your lip-synced video may not generate correctly if the audio and video durations are incompatible, CLI authentication is invalid, or media URLs are inaccessible. Ensure your RunComfy model routing matches your input format and quality requirements to avoid processing errors.