sipeter/CloneTTS

A lightweight, offline Android Text-to-Speech (TTS) engine enabling seamless system-wide voice cloning and high-fidelity text reading. / 运行在安卓本地的轻量级文字转语音 (TTS) 引擎，支持离线发音人提取、零门槛音色克隆与双擎系统级全局听书。

What it solves

CloneTTS is an offline, native Text-to-Speech (TTS) engine for Android that allows users to clone specific voices using very short audio samples. It eliminates the need for an internet connection and prevents data from being sent to external servers, providing a private, local alternative for reading books or long texts.

How it works

The engine uses zero-shot voice cloning acoustic models (via ZipVoice) and an efficient C++ inference pipeline (via sherpa-onnx). Users provide a 1-3 second audio clip and its exact transcript to create a custom voice. The app can be registered as a system-wide TTS engine or accessed via a local HTTP API for integration with external reader apps like Legado or Moon+ Reader.

Who it’s for

Android users who want high-quality, custom voice synthesis for reading apps, particularly those who prioritize privacy and offline functionality.

Highlights

  • Offline Voice Cloning: Create custom voices from just 1-3 seconds of speech.
  • NPU Acceleration: Supports hardware acceleration on flagship Qualcomm Snapdragon devices (8 Gen 2/3/Elite) for faster synthesis and lower heat.
  • System Integration: Works as a native Android TTS engine or via a local HTTP API for multi-role reading.
  • Customization: Includes pronunciation correction rules (plain text or regex) and adjustable sentence splitting for more natural pauses.
  • Performance Monitoring: Built-in benchmarking and real-time RTF (Real-Time Factor) monitoring to track inference latency.

Related

  • Project
    neuphonic/neutts

    A collection of open-source, on-device text-to-speech models that enable natural-sounding speech and instant voice cloning locally on mobile and embedded devices.

  • 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
    RVC-Boss/GPT-SoVITS

    A powerful few-shot voice conversion and text-to-speech system that can clone a voice using as little as 5 seconds of audio.

  • Project
    Xerophayze/TTS-Story

    A web-based multi-voice text-to-speech application for creating narrated stories and audiobooks using local and cloud TTS engines.

  • 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.