- Open Console › Content OS › Voices.
- On Install and health, choose Install next to Human voices. Weesify installs, in the background, the voices the library is made from, PyTorch for the processor, the Chatterbox Multilingual model (about 3.2 GB, checked against its checksums) and builds the ready-made house voice library. You can leave the page.
- When it finishes, open House voices to listen, rename voices, correct their description, switch any off, or offer a voice only on some plans.
More voices
- Tick I've checked its licence and accept it next to the LibriTTS-R voice, install it, then choose Build the library again to add about 32 more varied voices.
- Add a voice talent Weesify hired adds a voice from a 10 to 30 second clip, only after you confirm the talent signed consent. Their name and consent document stay private.
- Designed in workspaces lists voices people designed; Promote to house voice offers one to everyone.
If a step can't run
Open If an install step can't run here for the commands to paste in WHM › Terminal, then press Check again.