onyx-speechd - the Onyx Voice speech helper
Copyright (C) 2026 Integrators Resource. Licensed under the GNU General Public License v3 or later (LICENSE).

It is a separate program from Onyx Voice: the phone system starts it and exchanges text and audio files with it.
Its complete source is in this folder and is published with every release (onyxvoice.net/source/).

Components shipped with it, with their licences:
- sherpa-onnx (k2-fsa), Apache-2.0: https://github.com/k2-fsa/sherpa-onnx
  Its native library includes espeak-ng (GPL-3.0-or-later, https://github.com/espeak-ng/espeak-ng) and
  piper-phonemize (MIT, https://github.com/rhasspy/piper-phonemize). espeak-ng's source is available from that address
  and on request from Integrators Resource for three years from the release date.
- ONNX Runtime (Microsoft), MIT: https://github.com/microsoft/onnxruntime
- Kokoro-82M model and voices (hexgrad), Apache-2.0: https://huggingface.co/hexgrad/Kokoro-82M
  the full-precision v1.0 model from kokoro-onnx (thewh1teagle, MIT: https://github.com/thewh1teagle/kokoro-onnx)
  with sherpa-onnx's metadata added; voices, tokens, English lexicons and espeak-ng-data from sherpa-onnx's
  kokoro-int8-multi-lang-v1_0 package.
- .NET runtime (Microsoft), MIT: https://github.com/dotnet/runtime
