Lobe TTS is a high-quality, TypeScript-based toolkit for Text-to-Speech (TTS) and Speech-to-Text (STT) designed for both server-side and browser environments. It aims to provide developers with a robust and flexible solution for integrating voice capabilities into applications, addressing a perceived gap in readily available, high-quality frontend TTS libraries.

How It Works

The library offers distinct functionalities for server and browser use. Server-side, it supports multiple TTS engines including EdgeSpeechTTS, MicrosoftTTS, and OpenAI TTS, enabling voice generation with minimal code. For browser-based applications, it provides React Hooks and visual audio components that manage audio playback, timeline scrubbing, and visual styling. The architecture emphasizes ease of integration and a rich user experience for audio interactions.

Quick Start & Requirements

• Install: pnpm i @lobehub/tts or bun add @lobehub/tts
• Prerequisites: Node.js environment. For Node.js, WebSocket polyfill (import WebSocket from 'ws'; global.WebSocket = WebSocket;) is required. Next.js projects need transpilePackages: ['@lobehub/tts'] in next.config.js.
• Docs: https://tts.lobehub.com

Highlighted Details

• Supports server-side TTS with EdgeSpeechTTS, MicrosoftTTS, and OpenAI TTS.
• Provides React Hooks (useAudioPlayer) and visual components for browser-based audio playback.
• Offers extensive customization for audio track styling and timeline interaction.
• Designed with a focus on high-quality audio output and a smooth user experience.

Maintenance & Community

The project is actively maintained by LobeHub, which also develops related projects like Lobe Chat and Lobe Theme. Community contributions are welcomed via GitHub Issues.

Licensing & Compatibility

• License: MIT
• Compatibility: The library is ESM-only. Compatibility with Next.js SSR is addressed via transpilePackages.

Limitations & Caveats

The library is ESM-only, which may require specific build configurations for older module systems. While it supports multiple TTS backends, the quality and availability of voices depend on the chosen service.