Documentation for a newer release is available. View Latest

dengjen-tts

A cross-platform Rust engine for neural TTS models.

Features

  • Models: Piper, Kokoro, and MeloTTS ONNX voices, including Kokoro per-voice style embeddings

  • Phonemization: eSpeak-ng (100+ languages, IPA output) and Arabic diacritization via libtashkeel

  • Multi-speaker voices: select by speaker_id

  • Streaming synthesis: chunked output (chunk_size/chunk_padding) and a realtime gRPC stream

  • Prosody control: rate, pitch, and volume via dengjen-sonic-sys (libsonic)

  • Synthesis modes: lazy, parallel, and batched, selectable per request

  • Bindings: native Rust, C-API (libdengjen), Python (pyo3), gRPC (any language over the wire), and a CLI

Not yet supported, tracked on the issue tracker: native Go/Java/Kotlin bindings and a Matcha-TTS model loader. RHVoice-style formant/statistical synthesis is a different synthesis paradigm from this engine’s neural-ONNX pipeline and isn’t planned.

Crates

Crate Purpose

dengjen-espeak-phonemizer

Converts text to IPA phonemes using a patched version of eSpeak-ng

dengjen-model

Handles model loading and inference using onnxruntime via ort

dengjen-tts

Wraps DengjenModel and adds synthesized speech post-processing, including changing prosody. Also provides different modes of parallelism.

dengjen-tts-grpc

GRPC frontend for dengjen

libdengjen

C-API binding to dengjen

dengjen-tts-python

Python bindings to dengjen-tts using pyo3

dengjen-sonic-sys

Rust FFI bindings to Sonic: a C library for controlling rate, volume, and pitch of generated speech

License

Licensed under the GNU General Public License v3.0 or later (GPL-3.0-or-later). dengjen began as a fork of Sonata by Musharraf Omer, originally MIT-licensed; retained attribution is in the repo’s NOTICE file.

Want to help? Learn how to contribute to the ZirekHQ docs ›