Narration
Audiobooks and podcasts
Turn a single clean sample into hours of consistent narration with the pace, warmth and pronunciation of the original speaker.
Voice cloning studio
Clone a voice from ten seconds, then let it speak
Import a clean sample or record one in the browser. AI Voice Cloning mirrors the timbre, pace and accent of the speaker, then renders your script in that voice - with noise removal and accent optimization on hand when the source recording is not perfect.
Sign in with Google to open the studio. · Two voice slots are included with every account.
10-second reference32 output languagesNoise removal built inConsent-first voice slotsPreview before you render
Voice cloning studio
Clone a voice in one pass
Import a voice clip or record 10-60 seconds in the browser, tune the cleanup options, write your script and preview the cloned voice. Two voice slots are included with every account.
Engine voices-clone-v2 · New
AI Voice CloningVoice Cloning
Clone a voice using as little as a 10-second clean voice clip.
Import your Voice
or
Uploaded Samples
0/10Please upload a clean voice recording in a quiet environment. Reverb, multiple speakers, background noise, etc., will affect the voice cloning quality.
You can use the noise separation feature to reduce audio noise, but this may also impact the detail of the voice. Therefore, it is recommended to choose the best-quality segment when selecting the material for cloning.
Advanced Settings (Optional)
- Remove Background Noise
- Enable Accent Optimization
Text to Preview
151 / 300 charactersvoices-clone-v2
Voice slots remaining: 0/0
Generated voice results will appear here
Nothing is synthesized until you press Generate. Your request is held in the queue below.
Use cases
One cloning engine for every kind of spoken voice
Teams use AI Voice Cloning for narration, learning, marketing and support. The same studio covers the full production flow, from a ten-second reference to a finished voice track.
Learning
E-learning and courses
Keep every lesson in the same voice across updates and languages - no studio re-bookings when the script changes.
Marketing
Ads, demos and support
Produce voiceovers and product walkthroughs in the brand voice, then localise the same script for every market.
Why this studio
A clone that keeps the character of the voice
Most voice generators hand you a preset. Voice cloning keeps the speaker: the timbre, the pace, the small habits of pronunciation that make a voice recognisable.
Ten seconds is enough
A clean 10-60 second reference gives the model what it needs. Longer files are trimmed to the best-quality segment before cloning.
Cleanup before cloning
Noise removal and accent optimization run before the embedding is built, so phone recordings stay usable instead of failing the quality check.
Preview before you render
Write up to 300 characters, listen to the result and only then spend a render on the full script.
Consent-first voice slots
Every clone lives in your own voice library with a rights confirmation, and any slot can be deleted at any time to remove the embedding.
Example voices
Voices that fit the script
Start from a curated library of English voices, or clone your own and let the same studio read the script in that voice.
Calm Woman
Narration and audiobooks
Warm Narrator
Documentaries and explainers
Bright Host
Ads, social and promos
Deep Anchor
News, IVR and e-learning
How it works01 02 03
How to clone a voice
From a ten-second reference to a finished voice track in three passes.
Import or record
Add or drop a file up to 20 MB, or record 10-60 seconds in the browser. One speaker per sample keeps the clone stable.
Tune the cleanup
Toggle noise removal and accent optimization, then write up to 300 characters of script and preview the clone.
Render the clone
Press Generate: the studio builds the embedding and holds the render in the queue until a voice slot frees up.
More tools
More than one-way narration
The same workspace covers the rest of the voice workflow.
Voice Library
Every clone and built-in voice in one place, with the reference clip and consent record attached.
Voice Design
Describe the voice you want in words and design a new one when no recording exists yet.
Voice Isolator
Clean up interviews and field recordings and pull a single speaker out of the noise before cloning.
Frequently asked questions
What is AI voice cloning?
Voice cloning builds a model of a specific speaker from a recording of that speaker. The studio then renders any script in that voice, keeping the timbre, pace and accent of the original clip.
How long does the reference clip need to be?
Ten seconds is the minimum. The sweet spot is 10-60 seconds of clean speech from one speaker, with no music bed, reverb or overlapping voices.
Can I clone my own voice?
Yes. You can clone a voice only when you own it or hold written permission from the speaker. Every render requires the rights confirmation checkbox, and voice slots can be deleted at any time.
Which languages are supported?
The cloned voice keeps the accent of the reference speaker while the preview text stays English. Preview text is English-only for now, and the interface is English-only as well.
What happens to a noisy recording?
Turn on Remove Background Noise before cloning. The studio separates speech from room tone and hiss first, so noisy phone clips still produce a usable clone instead of failing the quality check.
Why is my render waiting in a queue?
Cloning runs on shared GPU capacity. When the cluster is full, your request keeps its place in the queue and starts automatically when a voice slot frees up.
Is there a free preview?
You can write up to 300 characters and preview the cloned voice before spending a full render. Accounts start with two voice slots so you can clone and test before subscribing.
How do I cancel or get a refund?
Subscriptions are billed through Stripe and can be cancelled at any time from the billing portal. The refund policy covers your first payment within 7 days.
Turn a ten-second sample into a voice you can reuse
Sign in with Google, import or record a reference clip, and preview the cloned voice in the studio.
Two voice slots included - no credit card required to preview.