How to Clone Your Voice
Regent clones your voice from a short audio sample and reuses it for every reel voiceover. A 15-second sample is enough to get started; for the best clone, upload 60–90 seconds of natural narration in the style you want your reels to sound.
Before you start
- Completed onboarding
- A quiet place to record, or an existing audio file (.mp3 / .wav)
Steps
- 1
Open Voice Cloning
Go to Reel Generation → Voice Cloning. Existing voice profiles, if any, are listed with their status.
- 2
Record or upload a sample
Record directly in the browser with the mic button, or upload an .mp3/.wav file. Minimum 15 seconds; 60–90 seconds of natural narration in your normal speaking style gives a noticeably better clone.
- 3
Label and configure
Name the profile, and choose Normal or Multilingual mode (pick the language for multilingual). Demo text is optional.
- 4
Clone
Click Clone my voice. Processing takes about two minutes. When the status shows Trained, a demo clip plays so you can hear the result.
- 5
Set the active voice
You can keep multiple profiles (for example one energetic, one calm) and switch the active one anytime. Reels use the active profile.
Frequently asked
How long should my voice sample be?+
15 seconds is the minimum and works. For the best quality, record 60–90 seconds of natural narration — talk the way you want your reels to sound.
Can Regent clone my voice in other languages?+
Yes. Choose Multilingual mode when creating the profile and select the language. English and Hinglish content styles are supported in onboarding.
Can someone clone a voice that is not theirs?+
No. Regent's consent policy requires you to clone only your own voice. See the AI Avatar and Voice Consent policy on the Security page.
Related guides
Public beta — 100 creator cap