- Stars
- 107
- Forks
- 15
- + today
- +2
- Created
- 5mo
Ranking data as of October 21, 2026 (UTC).
Star History
Today, hour by hour
Overview
Phonon TTS inference using xn It ranks #436 on GitTiger, gaining +2 stars on October 21, 2026 (UTC).
The project is written in Rust and has 15 forks. It was created 5mo ago.
What it is
`xn-ptts` is Phonon text-to-speech inference built on `xn`, from gradium-ai. The README preview positions Phonon as natural TTS that runs on phones, laptops, browsers and local devices rather than only on a server.
Capabilities explicitly described
- Two small model sizes, Phonon 31m and Phonon 40m, each described as small enough to ship. - Five languages: English, French, German, Spanish and Portuguese. - A custom inference stack that runs in browsers and on devices, described as fast across devices. - Voice choice: ready-to-use voices, or make your own. - Delivery control, with speaking rate set through a `durationdelta` parameter.
The project is written in Rust and licensed Apache-2.0. No topic tags are set and the classification confidence is 0.65, which reflects the sparse metadata rather than any doubt about the description.
Getting started
Model download links, build commands and API signatures are not included in the supplied evidence. Start from the repository README, which the preview indicates also has a "Try it" entry point, and check the Apache-2.0 terms if you intend to bundle the model.
git clone https://github.com/gradium-ai/xn-ptts.git
cd xn-ptts
# see README for setupFrequently Asked Questions
What model sizes are available?
Phonon 31m and Phonon 40m, both described as small enough to ship.
Which languages does it speak?
English, French, German, Spanish and Portuguese.
Can it run in a browser?
Yes. The README preview says a custom inference stack runs in browsers and on phones, laptops and local devices.
How is speaking rate controlled?
Through a `durationdelta` parameter, per the preview.