Build a Text-to-Speech (TTS) Converter in Python

Introduction

Text-to-Speech (TTS) is the technology that converts written text into spoken words. With Python, you can build a simple TTS converter using just a few lines of code — useful for accessibility tools, audiobooks, personal assistants, or even educational apps.

In this article, you'll learn:

• How TTS works

• How to install and use the pyttsx3 library

• How to convert any string or text file into audio

• How to customize the voice and speed

• How to save the output as an audio file

Step 1: Install the Required Library

We'll use pyttsx3, a Python library for offline, cross-platform text-to-speech.

Open your terminal or command prompt and run:

pip install pyttsx3

Unlike some TTS tools, pyttsx3 doesn’t need an internet connection — it uses your system’s built-in speech engine.

Step 2: Create the Python Script

Here’s the full code — copy, paste, and run it. Explanations are provided below.

import pyttsx3  # Import the text-to-speech engine
import os       # To interact with the file system

# Step 1: Initialize the TTS engine
engine = pyttsx3.init()

# Step 2: Set voice properties (optional)
voices = engine.getProperty('voices')
engine.setProperty('voice', voices[0].id)  # Change to voices[1].id for female voice if available
engine.setProperty('rate', 150)  # Speed of speech (default is ~200)

# Step 3: Define your text (optionally read from a file)
text = "Hello! This is a simple text-to-speech converter built with Python."

# Optional: Load text from a .txt file
# with open("sample.txt", "r") as file:
#     text = file.read()

# Step 4: Speak the text aloud
engine.say(text)
engine.runAndWait()

# Step 5: Save the speech to an audio file
engine.save_to_file(text, 'output_audio.mp3')
engine.runAndWait()

print("Speech playback complete and saved as 'output_audio.mp3'")

Step-by-Step Explanation

pyttsx3.init()

This initializes the text-to-speech engine based on your OS — using:

• SAPI5 on Windows

• NSSpeechSynthesizer on macOS

• espeak on Linux

Set Voice and Rate

voices = engine.getProperty('voices')
engine.setProperty('voice', voices[0].id)
engine.setProperty('rate', 150)

• voices[0] is usually the default male voice.

• You can try voices[1] for a female voice if available.

• rate adjusts the speaking speed — lower = slower.

Load Text from File (Optional)

with open("sample.txt", "r") as file:
    text = file.read()

This is useful if you want to read long passages, eBooks, or documents instead of hardcoding the string.

Speak the Text

engine.say(text)
engine.runAndWait()

This speaks the text out loud immediately using your system’s audio output.

Save to File

engine.save_to_file(text, 'output_audio.mp3')

Saves the speech as an MP3 (or WAV) file for later use — perfect for offline listening or sharing.

On some systems, the saved file format may be .wav. If .mp3 doesn’t work, try 'output.wav'.

Organizing Your Project

Here’s a simple project structure:

text_to_speech/
├── text_to_speech.py
├── sample.txt         # Optional text input
├── output_audio.mp3   # Saved output

Optional Enhancements

Use Cases

• Reading books or articles aloud

• Creating educational voice clips

• Making accessible apps for the visually impaired

• Generating voice alerts for automation systems

Final Testing Tips

Before using your script in a full project:

• Check that your microphone/speakers are working

• Ensure Python can access the TTS engine

• Try different voice settings if speech doesn’t sound right

Summary

With this script, you've built a fully functional TTS converter that can speak and save text as audio, giving you a strong foundation for advanced voice-based apps.

Happy Scripting ! !