Video EditingIntermediate

How to Clone Your Voice for Faceless Narration with ElevenLabs

Create a custom ElevenLabs voice clone so your faceless videos narrate in a consistent voice you own and control.

9 minIntermediate

Using the same stock TTS voice as thousands of other channels makes your faceless content blend in. Cloning your own voice gives every video a consistent, recognizable narrator that is genuinely yours. This guide covers recording a clean sample, creating the clone in ElevenLabs, and using it for narration. Only clone a voice you have the right to use, which is your own.

What you need

An ElevenLabs account with voice cloning (a paid tier)
A quiet room and a decent microphone
1 to 3 minutes of clean recorded speech (yours)
Your video scripts ready to narrate
About 20 minutes

Step 1: Record a clean voice sample

Record yourself reading naturally in the style you want for narration. Use a quiet room with no echo, keep a steady distance from the mic, and avoid background noise. One to three minutes of varied, clear speech is enough; quality matters far more than length.

Recording - sample checklist

[x] quiet room, no echo

[x] consistent mic distance

[x] natural narration tone

[x] 1-3 min, no background noise

[ ] export as WAV or high-bitrate MP3

Clean, consistent audio produces a better clone.

Step 2: Create the voice in ElevenLabs

In ElevenLabs, go to Voices, then Add Voice, and choose Instant Voice Cloning. Upload your sample, name the voice, and confirm you have the rights to it. The clone is ready in moments and appears in your voice list.

ElevenLabs - add voice

Add Voice > Instant Voice Cloning

Name: My Narrator

Samples: sample.wav (2:10)

[x] I confirm I own the rights to this voice

[ Add voice ]

Upload your sample and create an instant clone.

Step 3: Test and tune the settings

Generate a short test line with your new voice. Adjust Stability and Similarity: higher stability sounds more even but less expressive, lower stability sounds livelier but can wobble. Find a balance that sounds like you reading calmly.

Re-record if the clone sounds off

A clone is only as good as its sample. If it sounds robotic or unlike you, the fix is almost always a cleaner, more natural recording rather than tweaking sliders.

Step 4: Narrate your script via the API or app

Paste your video script into the ElevenLabs editor and generate, or call the API to batch many scripts at once. The API is handy if you produce videos in volume and want narration files generated automatically.

tts.sh

curl -X POST "https://api.elevenlabs.io/v1/text-to-speech/$VOICE_ID" \
  -H "xi-api-key: $ELEVENLABS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text":"Here is the trick broke people never hear about.","model_id":"eleven_multilingual_v2"}' \
  --output narration.mp3

Step 5: Drop the narration into your video

Import the generated MP3 into your editor as the narration track, then build the visuals and captions around it as you would with any faceless video. Every future video now uses the same recognizable voice.

Result: a custom voice clone you own that narrates your faceless videos consistently, generated by hand in the app or in bulk through the API.