Create and Use Voice AI Voice Cloning

Voice AI voice cloning allows you to create natural-sounding, on-brand Voice AI Agents by cloning a real human voice directly within your CRM. You can upload or record a short audio sample, select a target voice provider, and save the cloned voice to your library. Once the clone is ready, you can assign it to a Voice AI Agent using the standard Voice Selection interface.

Note: Voice AI - Clone Voice is currently a Labs feature. If you do not see this option, an administrator must enable it in Agency Settings under Labs.

What is Voice AI Voice Cloning?

Voice AI voice cloning transforms a real speaker's audio sample into a reusable custom voice, helping your agents sound more aligned with your brand. This enables a consistent calling experience without relying solely on stock voices or external cloning tools.

Key Benefits

• Branded, recognizable agents: Create a cloned voice that reflects your company, team, or preferred tone.
• Organized custom voice library: View cloned voices in a dedicated Cloned Voices section alongside Imported Voices.
• Clear processing status: Track the progress of your clone with visible status indicators.
• Recoverable failures: Failed clones remain visible with an option to retry after correcting the sample.
• Reusable agent voice option: Assign the cloned voice to any Voice AI Agent through the Voice Selection flow.

My Voices and the Cloned Voices Library

My Voices is the custom voice library used in Voice Selection for Voice AI Agents. With voice cloning enabled, cloned and imported voices are organized in the same area for easy management. Custom voices are divided into collapsible sections:

• Imported Voices: Custom voices imported from supported sources.
• Cloned Voices: Voices created from uploaded or recorded audio samples.

Note: There is a combined limit of 10 voices per company for both imported and cloned voices.

Audio Sample Requirements

The quality of the cloned voice depends on the source audio. A clean recording with one speaker yields the best results.

• Supported formats: .mp3, .wav, .m4a, .ogg, .webm, .mp4
• Maximum file size: 10 MB

Processing, Previewing, and Failed Clones

Cloned voices are not instantly ready after submission. They appear in the Cloned Voices section with a processing status. Once ready, you can preview and assign them to agents. If a clone fails, it remains visible with an option to retry. Common failure reasons include background noise, multiple speakers, low volume, echo, unsupported formats, or excessive file size.

How to Set Up Voice AI Voice Cloning

1. Open the target sub-account.
2. Navigate to AI Agents → Voice AI → Agent List.
3. Open or create a Voice AI Agent.
4. Click the agent's current voice to open Voice Selection.
5. Go to the My Voices tab.
6. Click Add Custom Voice.
7. Select Clone from the dropdown.
8. In the Clone Voice modal:
   • Enter a clear Voice Name.
   • Select a Target Voice Provider: ElevenLabs or Cartesia.
   • Add a Source Audio Sample by uploading a file or recording in your browser.
   • Check the Rights and Consent Confirmation box.
9. Click Save.
10. Wait for the cloned voice to appear under My Voices → Cloned Voices with a processing status.
11. Once ready, select the cloned voice, save the selection, and save the Voice AI Agent.
12. Use the Test Your Agent panel to verify the sound during a test call.

Frequently Asked Questions

Q: What providers can I use for cloned voices?
A: You can choose ElevenLabs or Cartesia as the target voice provider.

Q: What is the maximum upload file size?
A: The maximum file size is 10 MB.