VoiceStudio
VoiceStudio is an open-source, fully local app for voice cloning and design, video dubbing, transcription and audiobook creation. It clones a voice from a short clean recording, invents one from a description in words, and works with no account or API key.
Strengths
- Local, no account or API key: your voice and texts stay on your machine
- Video dubbing, transcription and audiobooks in the same app, where Voicebox stops at dictation and cloning
- A processor alone is enough, a graphics card speeds it up
Limitations
- The 646 languages are the project’s claim, quality not independently verified
- AGPL-3.0 licence for the application; models, voices and outputs have their own terms, which vary
- Only use cloning on voices you have permission for
- Without a graphics card, expect about 5 GB of disk and slower processing
Best for
- A creator who wants to dub a video or produce an audiobook without sending their voice to the cloud