Audiobook Studio 2.0 docsLocal Engine (XTTS Default)
XTTS is the default local engine: a private, on-machine voice-cloning engine that needs no account. It's the bundled example of a local engine.
At a glance
- The only engine installed by default.
- Runs locally, private and free to run.
- Clones a voice from samples into a reusable profile.
- A GPU makes it much faster.
Why a local engine is the default
XTTS is the engine that comes installed. It runs on your computer, free.
What XTTS is & GPU needs
XTTS is a neural voice-cloning engine. It runs on CPU but is dramatically faster on a CUDA GPU. Its model is downloaded on first use; see First Run.
The voice profile
XTTS turns your samples into a voice profile and reuses it every time it speaks.
Strengths & tuning
See Voice Cloning Quality for heuristics, and Engine Settings & Verification to configure and test it.