Chatterbox now powers our voice layer
We use Chatterbox, an open-source TTS family, for voice across Miransas products.
We now use Chatterbox as the open-source text-to-speech model behind our voice layer.
What changed
- Voice generation runs on our own infrastructure.
- Voice data is not sent to a third-party API.
- Every generated file carries a PerTh watermark.
Read the full model breakdown on the models page.
Next steps
| Area | Status |
|---|---|
| Multilingual support | In progress |
| Reference voice library | Planned |