Text to AI Speech Guide

A first reader for AI voices

Text to speech with AI, explained from the first word

Text to speech turns written words into spoken audio. AI voices learn how to do it from recordings of people talking, which is why they can sound natural and still trip over a name or a word like read.

Our story Contact

I read it then and will read it now.

Same spelling, two sounds. A voice has to work out from the sentence which one is meant.

From page to sound

How written words become speech

Many newer models learn these jobs together instead of running them one by one, but the jobs themselves stay the same.

  1. First, the text is tidied

    Abbreviations, dates and numbers are spelled out as the words a person would say.

  2. Then, words become sounds

    Each word is matched to its speech sounds, using the sentence around it to settle any doubt.

  3. Next, a rhythm is planned

    The model decides where stress falls, where the voice pauses, and where the pitch rises or drops.

  4. Last, the audio is made

    All of that is handed to a voice model, which generates the sound you hear.

On the page
Dr. Lee lives on Elm Dr.
Spelled out
Doctor Lee lives on Elm Drive.
As sounds
DOK-ter LEE livz on ELM DRYVE
With its rhythm
Doctor Lee lives on Elm Drive.

A short word list

Words worth knowing

The coral mark sits under the part of each word that takes the stress.

  • phoneme

    The smallest unit of sound that changes a word. Bat and pat differ by a single phoneme.

  • prosody

    The tune of speech: its stress, pitch and pacing. Flat prosody is what makes a voice sound robotic.

  • homograph

    Words spelled alike that sound different, such as the lead in a pipe and the lead in a race.

  • inflection

    The rise or fall of pitch within a word or phrase. It helps a listener hear a question as a question.

  • lexicon

    A list that tells a voice how to say particular words. Product names and surnames usually end up on one.

  • vocoder

    The part of a speech system that turns a planned sound shape into audio you can actually play.

Before you touch a slider

What changes how it sounds

The same sentence can come out warm, rushed or stiff. These are the controls that usually make the difference.

Let's eat, Grandma.

A short pause, and an invitation.

Let's eat Grandma.

No pause, and a very different dinner.

Many voices read a comma as a pause, so editing the text is often quicker than changing a setting.

Voice
The first choice and the largest. Age, accent and character decide whether the audio suits what it is reading.
Speed
Slower helps instructions and unfamiliar terms. Too fast and the words begin to run into each other.
Pitch
Small changes can lift or settle a voice. Large ones tend to make it sound processed.
Pauses
Breaks between sentences and paragraphs give a listener room to follow along.
Pronunciation
Respelling a word, or adding it to a custom list, fixes the names a voice keeps getting wrong.

Choosing one

Questions to ask of any tool

A tool that suits a weekly podcast may not suit a training course. Start with the job and work back.

  • What is the audio for?

    A short voiceover, a long course and a voice inside an app ask for quite different things.

  • Can you correct a word?

    Names and specialist terms are where voices slip, so check there is a way to fix them.

  • Where may you publish it?

    Licence terms decide whether audio can be used commercially, and free plans can differ from paid ones.

  • Whose voice is it?

    Cloning a voice needs that person's permission. See how a tool asks for consent and records it.

  • Where does your text go?

    Scripts are often processed on a provider's servers. Read the data terms before pasting anything private.

  • What do you get back?

    Look at file formats, audio quality and whether long scripts have to be split into pieces.

More to read soon

Articles on each of these subjects are on the way.

Our story Corrections policy

In their words

The questions this site is built around

  • How much better is AI text to speech than the built in system voices?
  • Which text to speech generator gives the most control over speed and emphasis?
  • Which text to speech AI handles punctuation and pauses most naturally?
  • Where can I find free text to audio conversion for a nonprofit newsletter?
  • Can you recommend free text to speech software for making study materials accessible?
  • Which text to speech tools sound closest to a real human narrator?
  • Are there text to speech free options that allow commercial use?
  • What AI voice text to speech platforms support multiple speakers in one script?
  • What text to speech online services work without creating an account?
  • Is there a text to audio free tool for converting long documents?