Clone a voice from audio samples (or invite a guest)

A voice clone becomes a character you can use as a host. Clone only a voice you have the right to use; see content ownership and rights for who owns it and what consent covers. To use a ready-made voice, see the voice library. If you already have a Professional Voice Clone elsewhere, import it from ElevenLabs. More options are in Characters and Voices.

Clone your voice

1. On the Characters and Voices page, click Create Voice Clone, then choose Clone your voice.
2. Upload or record samples, tick the box confirming you have the rights and consents to clone the voice, then click Next. If the samples fall short of the minimum or the box is unchecked, Jellypod marks what needs attention instead of continuing.
3. Jellypod clones the voice and drafts a name, language, and description from your samples.
4. On the Voice information step, play the preview and edit the drafted details. Click Finish Voice Clone.
5. On the confirmation screen, click Go to Voice Clone to open the character or Back Home to return.

If cloning fails, your samples stay in the dialog and the button reads Try again. If Jellypod cannot save the character after review, retrying reuses the voice it already cloned instead of cloning again.

To give an existing character a cloned voice, edit it and choose Voice Clone as the voice type. See managing characters.

Clones count against your plan: 2 on Starter and Educator, 8 on Creator, 20 on Business, and 100 on Enterprise (see plan limits). Archived characters do not use a slot, and pending guest invites do.

Audio requirements

Clear audio matters more than length: 1 to 2 minutes of clean speech is a good target.

Upload or record samples

Drag a file into the upload zone or click Browse files. Each sample uploads as soon as you add it, so you can keep adding while earlier ones finish.

Accepted formats: MP3, WAV, M4A, OGG, OGA, or FLAC audio, or a video file (MP4, M4V, MOV, WEBM, MKV, AVI). For a video, or an audio format Jellypod cannot use directly, Jellypod extracts the audio in your browser and converts it to MP3 (you see "Extracting audio from your file…"). If the file has no audio track or cannot be processed, you get an error and can pick another file.

Click Record Audio to record in your browser instead, then click Start and Stop Recording when finished. The recording is added to the Audio samples list.

Each sample appears in the Audio samples list with its file name and duration, plus buttons to play, download, and remove it. A row shows Uploading… while in progress, or Upload failed with a Retry button. Removing a sample cancels its upload. Continue once every upload finishes; retry or remove failed ones first.

For a better clone, introduce yourself and say what you do (Jellypod uses it to build the character), record somewhere quiet, and vary your tone and pace. Background noise and room echo become part of the clone.

Invite a guest to clone their voice

A guest can record a clone into your account without a Jellypod account.

1. Click Create Voice Clone, then choose Invite a guest. You can also click Invite a Guest on the Voice Clone Invites page (/characters/invites).
2. Enter a Character Name and a Backstory, then click Send invite.
3. Copy the invite link to share it yourself, or enter your guest's email address and click Send email.

The guest opens the link, uploads or records samples that meet the same requirements above (each with an inline Retry if it fails), and submits. The character then appears on the Characters page as a regular Voice Clone character and you get an email.

An invite link works once and expires after 24 hours. From the Voice Clone Invites page you can copy the link, send the email, or Delete an invite. If a link expires, create a new invite. While an invite is pending or unexpired it counts toward your voice clone limit, so delete unused invites if you hit it.

If your clone doesn't sound right

1. Check the recording. Noise, echo, and a distant microphone carry into the clone.
2. Re-record. Use a clear, consistent tone with some variation, not a monotone read, and upload the new samples.
3. Change the Voice style. In the character's voice settings, More Consistent stays closest to the recorded voice. More Expressive is more emotive and supports audio tags like [whispers].
4. Regenerate the speech blocks. If the clone works overall but specific lines sound off, regenerate those blocks for a different rendering with the same clone.

If results stay poor, start over with new, higher-quality samples.

Open Help & Support in the Studio sidebar to chat with the team, or email support@jellypod.com.