AI Speakers overview

Create and edit content with your own voice clone or one of Descript's AI stock voices. an AI model of your voice or one from our ultra-realistic AI stock voices. This article will introduce you to all the things you can do with AI Speakers in Descript.

Create a custom AI Speaker

Screenshot of Descripts training and authorization statement

Create an AI Speaker in your own voice with 30 seconds of audio. You'll be editing and generating audio with your AI voice in under a minute. Learn how to create a custom AI Speaker

Generate text-to-speech

GIF showing a script being written in Descript using text-to-speech

Descript's AI Speakers let you create a text-to-speech model of your voice or select one from our ultra-realistic stock voices. Learn how to generate text-to-speech

Changing words or phrases with Regenerate

2024-11-11_13-43-47-2.gif

Use a custom AI speaker and Descript's Regenerate feature to address these hiccups without the hassle of re-recording. Learn how to Regenerate audio in a project

Not available for stock voices

Our stock voices cannot be used when regenerating existing audio. To dub or correct existing audio, you'll first need to create a custom AI speaker.

Fixing tone or mispronounced words with Regenerate

2024-11-11_13-43-47.gif

Correct mismatched tone, mispronounced words, enhance lackluster dialogue, or remove annoying background noise from your recordings. Regenerate makes impossible edits possible — with one click. Learn how to use Regenerate

Adjust and fine tune AI audio

Convert AI voice clips into audio 

AI generated audio works differently than recorded audio in the script. You can't adjust the word boundaries speed, or use a Timeline tool on sections of AI audio. You'll need to first convert the AI voice clip into audio that you can edit.  
  1. Hover over an AI audio clip in the Timeline at the bottom of your editor.
  2. Select convert to audio.
If you're only seeing the wordbar, you'll need to expand the Timeline.

ADD ALT TEXT HERE

Change clip speed

After converting an AI voice clip to audio, you can speed up or slow down the clips from the sidebar.

  1. In the timeline, select the clip you want to adjust.
  2. In the playback section of the Layer Panel, click and drag the speed value or type in a value.

Screenshot: Properties panel with red arrow pointing where to adjust clip speed

These adjustments are different from adjusting your composition's playback speed:

  • Clip speed changes will be included when you export your composition.
  • Playback speed does not affect your exported content and only changes the speed during playback within the Descript project.

Add fades

After converting AI voice clip to audio, You can quickly create audio fades and crossfades by dragging a transition handle in the Timeline.

creating_crossfades_and_fades_V60.gif

You can also adjust and apply fades or crossfades by clicking on the transition handle of your clip.

Adjusting_fades_and_crossfades_V60.png

Adjusting the tone and style of your AI Speaker

If you've already created a custom AI speaker but want a different tone or style, try making a new one with that specific delivery in mind. When recording your new voice sample, speak clearly and intentionally in the tone you want the AI to capture, whether that's casual and friendly, formal and authoritative, or anything else. How you speak during the recording directly shapes how your AI speaker will sound.