Skip to main content
Skip to workspace

AI Audio Tools

Loading generation options…

Listen to a voice, then try your script

Calm WomanOfficial voice sample
—
Trustworthy ManOfficial voice sample
—

Voice previews only. Choose your voice separately in the form.

About the voice previews

Official voice samples use different text; generation version is not specified. Portraits are illustrative, not the speakers' identities. Choose a compatible voice separately in the form.

Input example

A clear introduction

Example script

Welcome back to the studio. Today, we are starting with a simple idea: a short melody, a few words and a quiet place to listen. We will build one version, hear what works, and keep the details that make the song feel like our own.

About this example
What you provide
  • Your spoken text.
  • A voice selected in the form.
What to expect

A spoken recording of your script with the selected voice and supported settings.

What to check
  • Listen for names, pronunciation and pauses.
  • Keep delivery instructions separate from words that should be spoken.

Input example

A scene narration

Example script

The last light settled across the water. A boat moved slowly toward the far shore, leaving a narrow line behind it. For a moment, the whole city seemed to pause and listen. Then the first notes of a piano drifted out through an open window.

About this example
What you provide
  • A narration script.
  • A compatible preset or owned voice.
What to expect

Narration to bring into an audio or video editor after review.

What to check
  • Compare the pace with the scene rather than assuming the preset sample sets the duration.

Input example

A short announcement

Example script

Our new track is here. Put your headphones on, turn it up, and listen for the moment the chorus opens. You can find the full song and the story behind it on our release page. Tell us which part you want to hear live.

About this example
What you provide
  • The exact announcement text.
  • Your selected voice and supported delivery settings.
What to expect

A spoken announcement, separate from any song-generation task.

What to check
  • Check the emphasis, music title and any call to action before publishing.
About these examples

Read a complete script and compare official preset voice samples. The samples use different text and do not demonstrate these scripts or a specific generation version.

A CLOSER LOOK

AI Audio Tools

Create speech and dialogue, clean up a voice recording, or generate a short sound effect. Choose the audio workflow that fits your project.

Choose the Speech Workflow for Your Script

Your taskChooseWhat you prepare
One narratorText to SpeechThe script and an available compatible voice
A conversationText to DialogueOrdered speaker turns and role-to-voice assignments
A reference voice for this recordingReference VoiceThe current required sample/consent pair and script, or a supported preset
A singing voice on a songThe separate music singing-voice workflowA song and a ready compatible singing model
A microphone, pop shield, headphones and script notebook in a quiet recording space.
AI-created editorial illustration.
01

Text to speech starts with a readable script.

Write the words as you want them spoken. Short sentences, useful punctuation and clear phrasing give you a better basis for reviewing the audio. Say names, numbers and abbreviations aloud before you submit. A script that reads well on a page may still need small changes when you hear it.

For text to speech, add the script and review punctuation, names and numbers. Choose an available voice, listen to its preview and check the selected settings and credit cost. Listen to the generated speech, edit wording when needed and download the version you want to keep. Dialogue, reference voices and cloning have different input preparation steps, so choose the matching function before preparing a recording.

An adult narrator considers vocal delivery in a quiet studio with headphones and a simple script.
AI-created editorial illustration.
02

Choose a voice for the message.

Listen to the available voice previews and choose one that suits the script: a calm explanation calls for a different delivery from an energetic introduction. Use the options the selected model offers. A voice preview helps with direction, but your exact wording can produce a different rhythm or emphasis.

A narrator listens back to a spoken recording while checking printed script pages with a pencil.
AI-created editorial illustration.
03

Review narration before you use it.

Listen all the way through, paying attention to pronunciation, pauses and sentence endings. Revise the text when a phrase sounds crowded or unclear, then create another version if needed. Keep the spoken track separate from any music you plan to add so you can review the balance in your editing workflow.

How it works

  1. 01

    Choose the audio workflow

    Choose speech, dialogue or a reference voice for a script; Clean voice for a recording; Sound effects for a short sound described in words.

  2. 02

    Prepare the matching inputs

    Review the script and voice for speech, assign turns for dialogue, or prepare the required clean recordings and consent material for a reference voice. Check the selected model, input requirements and displayed cost.

  3. 03

    Check the delivered audio or saved voice

    Listen to the complete generated speech before downloading it. For a saved voice, check its ready state and compatible use before choosing it in another task; a voice setup is not itself a finished narration.

Questions before you create

Can a text-to-speech voice sing my song?

Spoken narration and singing are different tasks. Use Create Music or Add Vocals for supported sung-music options.

Are all languages and voice styles supported?

Availability depends on the selected model and voice. Check the tool options and test the actual wording you need.

Where are vocal removal and audio tools?

Choose Separate vocals for vocals and accompaniment, Split stems for available instrument parts, or Export WAV for a supported song export. These source-song tools are distinct from the speech and voice workflows introduced here.

Create spoken audio