Voice credits

The people behind the voices.

Scriber’s natural narration uses open-source software and openly licensed model weights.

Kokoro-82M by hexgrad — Apache 2.0 model weights and voices.
Transformers.js 3.8.1 — Apache 2.0. ONNX Runtime — MIT.
HeadTTS pronunciation modules — MIT, using the CMU Pronouncing Dictionary under its BSD-style license. Scriber adapts the dictionary loader for its removable local cache.
KokoroSwift 1.0.11 and MLX Swift — MIT. MisakiSwift and MLXUtilsLibrary — Apache 2.0.
KokoroTestApp supplies the native model and voice files — Apache 2.0.
LameJS / @breezystack/lamejs 1.2.7 encodes MP3 audiobooks on your device — GNU LGPL v3. Source and notices.

Read the complete voice software notices. The linked projects provide their licenses and source. These credits concern the voice software and model, not ownership of your writing.