How to Clone Your Own Voice with Instant Voice Cloning

Instant Voice Cloning builds a reusable voice from a sample of your own speech. Once created, that voice appears in the Text to Speech picker like any prebuilt voice, so you can generate narration that sounds like you without recording every line by hand. This guide covers preparing a clean sample and creating the clone.

What you need

• A paid ElevenLabs plan (Instant cloning requires at least the Starter tier)
• One to three minutes of clean recorded speech, ideally a single speaker with no background noise
• A quiet room and any basic mic, even a phone, for recording
• Legal right to clone the voice (use your own voice or one you have permission to use)

Step 1: Record a clean sample

Record yourself reading naturally for about a minute. Avoid music, echo, and other voices. Speak the way you want the clone to sound, since the model copies your tone and pacing, not just your timbre. Export the recording as a WAV or MP3.

Step 2: Open the Voices library

In the sidebar click Voices, then Add a new voice, and choose Instant Voice Clone.

Step 3: Upload and label

Drag your sample file into the upload area, give the voice a clear name, and confirm the verification checkbox stating you have the right to clone this voice. You can add a short description to remind yourself which project it belongs to.

Step 4: Create the voice

Click Add voice. Cloning is fast and usually finishes in well under a minute. Your new voice now appears at the top of the Voices list and inside the Text to Speech voice dropdown.

Step 5: Test with a short script

Open Text to Speech, select your cloned voice, and generate a sentence you did not include in the sample. Listen for accuracy. If the clone sounds flat, your sample may have been too monotone; record a livelier sample and create a fresh clone.

Result: you can now type any script and hear it read back in your own voice, which is ideal for batching narration when you do not have time to record every take.