thewh1teagle/kokoro-onnx

TTS with kokoro and onnx runtime

What it solves

It provides a lightweight, high-performance way to run the Kokoro-TTS text-to-speech system using the ONNX Runtime, making it easier to deploy and run locally without needing the full original model environment.

How it works

The project converts the Kokoro-TTS model to the ONNX format, allowing it to run on both CPU and GPU. It utilizes the kokoro-onnx Python package and requires a model file (.onnx) and a voice binary file (.bin) to generate audio from text.

Who it’s for

Developers who want to integrate high-quality, multi-language text-to-speech capabilities into their applications with low latency and a small memory footprint.

Highlights

  • Multi-language support: Capable of handling various languages.
  • High performance: Achieves near real-time performance on macOS M1 chips.
  • Lightweight: The model is approximately 300MB, with a quantized version available at around 80MB.
  • Crosspatform: Supports both CPU and GPU execution via ONNX Runtime.

Related

  • Project
    KittenML/KittenTTS

    A lightweight, ONNX-based text-to-speech library that provides high-quality voice synthesis on CPU with models as small as 25MB.

  • Project
    OpenMOSS/MOSS-TTS-Nano

    A multilingual tiny speech generation model with 0.1B parameters designed for real-time, CPU-based inference and high-fidelity audio output.

  • Project
    PierrunoYT/Kokoro-TTS-Local

    A local implementation of the Kokoro-82M text-to-speech model featuring a web interface, 54 multilingual voices, and full offline support.

  • Project
    tronghieuit/v-tts

    A lightweight Vietnamese text-to-speech system featuring multi-speaker synthesis and zero-shot voice cloning that runs efficiently on CPU.

  • Project
    samuel-vitorino/sopro

    A lightweight 120M-parameter text-to-speech model family that enables high-quality voice cloning and streaming audio generation on laptop CPUs or in the browser.