Summary

CloneTTS is a lightweight, offline Android Text-to-Speech (TTS) engine designed for system-wide voice cloning and high-fidelity text reading. It empowers users to create custom voices from short audio samples and integrate them seamlessly into Android applications, all processed locally without an internet connection.

How It Works

The engine operates entirely offline on Android devices, enabling users to clone voices by providing just 1-3 seconds of reference audio. It registers as a system TTS engine, allowing compatibility with various reading applications, and also offers an HTTP API for advanced integration. This local-first approach ensures user privacy and enables functionality without network access.

Quick Start & Requirements

• Installation: Download the latest .apk from the GitHub Releases page.
• Prerequisites: Android device. Crucially, users must configure battery optimization settings to "Unrestricted" or "Allow background activity" and lock the app in the recent tasks view to prevent system termination.
• First Launch: Requires a few seconds for background model data extraction.
• Documentation: GitHub Repository

Highlighted Details

• Rapid Voice Cloning: Generates custom voices from as little as 1-3 seconds of recorded speech.
• System-Wide Integration: Functions as an Android system TTS engine, compatible with popular readers like Moon+ Reader and Legado.
• Dual Integration Modes: Supports seamless integration via the Android system TTS framework or a local HTTP API server for advanced use cases.
• Advanced Customization: Features include multi-voice rotation, adjustable speech rate/volume, custom pronunciation rules (text/regex), and importable polyphone dictionaries.

Maintenance & Community

• Project Homepage: https://github.com/sipeter/CloneTTS
• Community: Official QQ Group: 640190207.

Licensing & Compatibility

The README does not specify a software license. However, it includes a disclaimer emphasizing that unauthorized extraction or use of others' voices is strictly prohibited and users bear responsibility for misuse, suggesting a focus on personal, non-commercial use and potentially restrictive terms.

Limitations & Caveats

The application's background processes are highly susceptible to system-level battery optimization and task management, requiring mandatory, non-default user configuration to ensure stable operation during extended reading sessions. The absence of a formal license and explicit usage restrictions necessitates careful consideration for any integration beyond personal use.