v1.0.0

VocaVid 0.1.0 – First public release

VocaVid is a local, storyboard-based workflow for AI music videos using ComfyUI.

This first public release covers the entire process:

• Create a project with WAV files, lyrics, style, and references
• Analyze lyrics using Whisper and automatically segment them
• Manually correct timing and segments as needed
• Generate a scene plan as well as image and motion prompts
• Generate images, avatar variants, and video clips
• Review, approve, and selectively re-render clips
• Export a Kdenlive project and the final MP4
• Analyze reel candidates and export them vertically

Important

• VocaVid runs locally and requires Python, ComfyUI, FFmpeg, and a CUDA-capable GPU.
• Models and custom nodes depend on the workflow used. The included workflow JSONs serve as the authoritative reference.
• The project is still under active development. Feedback, bug reports, and suggestions for improvement are welcome.

More information and documentation:
https://vocavid.de/