rusty_tts

A unified FastAPI-based text-to-speech service that supports multiple TTS engines and providers, allowing you to generate audio from text using various synthesis methods.

Features

• Multiple TTS Providers: Support for cloud-based APIs and local TTS engines
• Voice Selection: Different voices available per provider
• Audio Caching: Generated audio files are cached to avoid regeneration
• RESTful API: Simple HTTP endpoints for easy integration
• Streaming Audio: Direct audio file streaming for web playback
• Provider Auto-Detection: Automatically detects which TTS engines are available

Supported Providers

The API supports various TTS providers including:

• Cloud-based services (like Pollinations)
• Local TTS engines (eSpeak, Festival, Flite, DECtalk, SAM)
• AI models (Coqui TTS)
• Custom effect engines ((currently broken) LPC-style processing)

Quick Start

1. Install dependencies:

   pip install fastapi requests uvicorn

2. Run the server:

   python app.py

3. Generate speech:

   GET /tts?text=Hello+World&provider=pollinations&voice=alloy
   

API Endpoints

• GET / - API information and available providers
• GET /tts - Generate audio from text
• GET /play/{filename} - Stream audio files
• GET /providers - List all available providers and voices
• GET /files - List generated audio files
• GET /health - Health check