edge-tts

edge-tts: Microsoft Edge Online TTS Integration for Python

edge-tts is an open-source Python library designed to harness the power of Microsoft Edge's online text-to-speech (TTS) service. This service, powered by advanced neural voices from Microsoft, allows developers and users to convert text into high-quality, natural-sounding speech without the need for installing Microsoft Edge browser, a Windows operating system, or obtaining any API keys. It's particularly useful for applications requiring multilingual voice synthesis, such as accessibility tools, virtual assistants, audiobooks, or automated content narration.

Key Features

• Seamless Integration: Use it directly in Python scripts or via simple command-line interfaces (edge-tts for generation and edge-playback for playback with subtitles).
• No Dependencies on Edge or Windows: Runs on any platform (Linux, macOS, Windows) as long as you have Python and internet access.
• Multilingual Support: Access a vast library of voices across numerous languages and accents, including neural voices with personalities like friendly or positive tones.
• Customization Options: Adjust speech rate, volume, pitch, and select specific voices to tailor the output.
• Output Flexibility: Generate audio files (e.g., MP3) and synchronized subtitles (SRT) for easy integration into videos or apps.
• Command-Line Convenience: Quick TTS generation and playback without writing code.

Installation and Setup

To get started, install via pip:

pip install edge-tts

For command-line only use, prefer pipx:

pipx install edge-tts

Note that edge-playback requires the mpv player (except on Windows) for immediate audio playback with subtitles.

Basic Usage Examples

Command-Line TTS Generation

Generate a simple audio file:

edge-tts --text "Hello, world!" --write-media hello.mp3 --write-subtitles hello.srt

Play back with subtitles:

edge-playback --text "Hello, world!"

Voice Selection

List available voices:

edge-tts --list-voices

This outputs a table of voices like af-ZA-AdriNeural (Female, General, Friendly). Use a specific voice:

edge-tts --voice ar-EG-SalmaNeural --text "مرحبا كيف حالك؟" --write-media hello_ar.mp3 --write-subtitles hello_ar.srt

Audio Customization

Modify speech attributes:

• Slow down rate: --rate=-50%
• Lower volume: --volume=-50%
• Adjust pitch: --pitch=-50Hz Example:

edge-tts --rate=-50% --text "Hello, world!" --write-media slow_hello.mp3 --write-subtitles slow_hello.srt

Python Module Usage

For programmatic access, import and use the library. The project provides examples in /examples/ and utility functions in /src/edge_tts/util.py. Key steps include:

1. Listing voices: edge_tts.list_voices()
2. Creating a communicator: communicate(text, voice)
3. Saving audio: Stream to a file.

Example code snippet:

import asyncio
import edge_tts
 
async def tts_example():
    text = "Hello from edge-tts!"
    voice = "en-US-AriaNeural"
    communicate = edge_tts.Communicate(text, voice)
    await communicate.save("output.mp3")
 
asyncio.run(tts_example())

Limitations and Notes

• SSML Support: Custom SSML is limited to what Microsoft Edge generates (single <voice> and <prosody> tags). Advanced SSML is not supported due to service restrictions.
• Internet Required: Relies on Microsoft's online service, so an active connection is needed.
• Voice Availability: Voices are subject to Microsoft's updates; use --list-voices for the latest.

Community and Integrations

With over 9,500 GitHub stars, edge-tts is widely adopted. It's integrated into projects like:

• hass-edge-tts for Home Assistant.
• Podcastfy for podcast generation.
• tts-samples for voice testing.

This library democratizes access to premium TTS technology, making it ideal for AI-driven applications, educational tools, and content creators seeking cost-effective speech synthesis.