Clone your own voice

Record your own voice so your AI agent can speak on calls in it.

Where to find it

Voice › AI voices

Open this screen
Clone my voice card with the consent step and the reading steps

Steps

  1. Open Voice › AI voices and press Clone my voice.
  2. Under 1. Your consent, enter Your name, choose Language you will read in, and tick This is my own voice. I will not clone anyone else's voice.
  3. Press Record the sentence above, read the consent sentence aloud, then press Stop.
  4. Under 2. Read the text aloud, press Record for each paragraph and Stop when you finish it. Use Record again for any paragraph you want to redo.
  5. Under 3. Create the voice, leave Provider on Automatic (xAI first, then ElevenLabs) or pick one, then press Create my voice.
  6. When it is ready, press Listen to a sample, then use it for your agent (see Use your cloned voice).

Settings on this screen

SettingWhat it doesDefault
Your nameThe name read in the consent sentence.—
Language you will read inJapanese or English text to read.日本語
ProviderWhich service makes the voice.Automatic (xAI first, then ElevenLabs)

Troubleshooting

"Too short. Read the whole text, then stop."

Read every paragraph to the end before pressing Stop. The whole reading needs at least 20 seconds.

"Too much background noise. Find a quieter place."

Record again in a quiet room, in your normal phone voice.

A new clone shows Waiting for review.

Someone at the service listens to the consent recording first. Until then the voice cannot speak on calls, be previewed or make a test call.

Tips

  • Read in your normal phone voice, in a quiet room.
  • The meter under the counter shows how many paragraphs are recorded.

Good to know

  • Only clone your own voice.
  • An organisation can keep 3 cloned voices; delete one to make a new one.
  • Only workspace managers can clone, and the browser must allow the microphone.

Related how-tos

Start free Home