Voice Cloning — Clone Your Voice for Your AI Agent

Written By Maruf Ansari

Last updated 2 days ago

Give your AI Agent a voice based on your own recording instead of choosing one from the Voice Library. This guide covers uploading an audio file or recording in the portal, adding voice details and selecting the cloned voice for your Agent.

Clone Your Voice Overview

Clone Voice allows Business users to create a custom AI Agent voice using an audio sample and select it in the Agent's Identity settings. You can upload an existing recording or record a new sample, provide the required voice details, and submit it for cloning. Once the voice is available, you can select it for your AI Agent and perform a controlled test to check the audio quality. Using a clear recording with minimal background noise, echo, or distortion helps achieve better voice quality.

Important notes

The Upload guidance recommends at least 1 minute of audio and says to avoid more than 3 minutes.
Creating a voice and selecting it for an Agent are separate tasks. Verify the selected voice before relying on it for live calls.

Before you start

Open the Agent you want to configure in the Business Portal.
For Upload, prepare an MP3 or WAV audio sample. Review the audio guidelines below.
For Record, have a working microphone and a quiet place to speak.
Use your own voice. Avoid including private Business information or other people's conversations in the sample.

Open Clone Your Voice

In your Agent, open Configure → Identity.

Under Select Voice, choose Clone Voice, then select Add New.

The Clone Your Voice dialog offers Upload and Record. Choose one method below.

Option 1: Upload an audio sample

Select Upload in Clone Your Voice.

Drag your audio file into the upload area, or select Select file and browse to the file.

In Add Voice Details, enter a recognizable Voice Name and select Language and Gender.

Audio guidelines for uploads

Guideline	What the dialog specifies
Accepted formats	MP3 (preferred) or WAV.
Recording length	At least 1 minute of audio; avoid more than 3 minutes.
Audio quality	Clear audio without reverb, artifacts or background noise. Reverb is lingering echo; artifacts are unwanted audio distortions.
Audio codec	MP3 at 128 kbps or above is recommended. This refers to the audio encoding bitrate.

Option 2: Record your voice

Select Record in Clone Your Voice.

Select Start Recording. Speak about anything, such as Talk about your favorite city., or read the passage under Or Read along.

Watch the waveform and elapsed timer. Select Restart if you need another take, or Finish to complete the recording.

In Add Voice Details, enter a recognizable Voice Name and select Language and Gender.

Select Submit.

Configuration reference

Field or Option	Purpose	Example or Limitation
Select Voice → Clone Voice	Choose a cloned voice instead of Voice Library.	Voice Library is the alternative.
Add New	Open Clone Your Voice.	Not needed to select an existing cloned voice.
Upload / Record	Choose how to provide the audio sample.	Upload a file or record using a microphone.
Voice Name	Identify the voice.	For example, Front Desk Voice; naming limits are not shown.
Language	Provide the voice's language detail.	The available choices are not shown.
Gender	Provide the voice's gender detail.	The available choices are not shown.
Voice	Select an existing cloned voice in Identity.	Choose the intended entry from the dropdown.

Select the voice and verify the result

Return to Configure → Identity and review Select Voice → Clone Voice.
Open the Voice dropdown. Look for the name you submitted and select it when available.
Reopen Identity and check that the intended voice remains selected.
Run a controlled voice test with the Agent before using the change for live calls. Listen to the greeting and a short exchange for clarity and the intended voice.

Expected result: The intended cloned voice is visible and selected in the Voice dropdown, and your test confirms the Agent uses the intended voice.

Troubleshooting

Problem or Symptom	Likely Cause	Solution
Clone Voice or Add New is unavailable.	Access or availability is not established by the screenshots.	Confirm you are in the intended Agent's Identity settings, then ask your Agency support team. Product validation required: Confirm access requirements.
Your audio file is not accepted.	Its format may not match the displayed MP3 or WAV guidance; exact rejection rules are not shown.	Check the format and audio guidelines, then retry with a suitable file.
The recorded sample contains noise or echo.	The recording environment may not match the clear-audio guidance.	Move to a quieter space and use Restart for another take before selecting Finish.
Recording does not capture your voice.	Microphone access or the selected microphone may need checking.	Check your browser's microphone access and device.
The new voice is missing after Submit.	Processing and failure states are not shown.	Check the Voice dropdown and contact your Agency support team with the voice name and what happened.
The Agent still uses a different voice.	The wrong voice may be selected, or the change may not yet be applied.	Recheck Clone Voice and the Voice dropdown, then repeat the controlled test.

Frequently Asked Questions

Can I upload audio instead of recording in the portal?

Yes. Clone Your Voice offers both Upload and Record. Upload accepts MP3 (preferred) or WAV.

How long should my uploaded sample be?

At least 1 minute and to avoid more than 3 minutes. It also emphasizes clear audio over total length. These recommendations do not establish enforced limits.

Which voice details are required?

Voice Name, Language and Gender are marked required in Add Voice Details. Complete them before selecting Submit.

Does submitting a cloned voice automatically assign it to my Agent?

Do not assume it does. Return to Identity, choose Clone Voice and check the Voice dropdown for the intended selection.

Was this helpful?

Still need help? Ask the team