What Is Text to Speech?

A text-to-speech system works in stages. First it cleans the text up, expanding numbers, dates and abbreviations into words. Then it works out how each word should be pronounced and how the sentence should be paced and stressed. Finally it generates the audio itself, as a waveform you can play or save as a file such as MP3 or WAV.

• Built-in device voices come with your phone or computer. They are free and instant, and their quality depends on the device.
• Cloud voices run on a provider's servers and are usually the most natural-sounding. Most modern ones use neural networks (see what is neural text to speech).
• Older synthetic voices stitched together recorded fragments or used rule-based models. They tend to sound flatter and more robotic.

Common uses are voiceovers for videos and courses, podcasts and audio versions of articles, phone greetings and menus, accessibility for people who prefer to listen or who cannot read a screen easily, and language learning. It is different from speech to text, which goes the other way and turns spoken audio into written text.

GetMeAudio lets you type or paste a script, place exact pauses, add background music, and choose from 938 voices across 153 languages and accents. You can draft for free with a voice built into your device, preview the real cloud voice, and pay only when you download the finished MP3 or WAV.