phonon-31m-en-coreml

Core ML build of Phonon 31m (English text-to-speech) by Gradium, taken unchanged from the v0.4.0 release asset phonon-31m-en-coreml.zip of https://github.com/gradium-ai/xn-ptts (zip SHA-256 10f3a99fd39256db44f826267ef21d78cca959b16c0554926344c09fde27b436), with the voices reduced to nine (default, Harper, Hadley, Miles, Preston, Maeve, Bryony, Edwin, Hugh) and bundle.json listing exactly the files present. Model: https://huggingface.co/gradium/phonon-31m-en (Apache-2.0). Voices: https://huggingface.co/gradium/phonon-voices (Apache-2.0).

Licensed under the Apache License 2.0. Phonon is by Gradium, built on Kyutai's Pocket TTS.

For iOS 18 / macOS 15 with the ptts Swift package (PhononModels.download(from:progress:), which checks every file against bundle.json). This copy is not made or endorsed by Gradium or Kyutai.