AudioPod AI
  • Pricing

Voice Studio

Cloning a voice

Whose voice you may clone and why that comes first, what makes a reference recording good, how many clones each plan holds, and the cases where a clone is the wrong tool.

Lesson 3 · core · 8 min read

Open Voice Studio

Careful

You may only clone a voice you have the right to use — your own, or someone else's with their explicit, informed permission. Nothing about the upload can verify whose voice it is, so this is entirely on you.

That callout is at the top rather than the bottom because it is the part of this lesson that matters to someone other than you. Everything else here is about getting a good result; this is about not wronging a person who is not in the room. Nothing in the upload can tell whose voice a recording is, so the check does not exist in the software — it exists in you.

Consent, in the cases that come up

  • Your own voice needs no one else's permission, and it is the only case that is simple.
  • Someone else's voice needs their permission — specifically for a synthetic clone, not merely permission to have recorded them. Consent to be interviewed is not consent to be cloned.
  • A public figure, a performer or a character from a film is somebody's voice and usually somebody's livelihood. That it is easy to find a clean recording of them is not a permission.
  • A voice you bought a recording of is licensed for the recording, not for the voice. Stock audio, an audiobook you own and a podcast episode are all the same case here.
  • Permission is worth having in writing, because the person who needs it later is you.

Note

The distinction people miss is between permission to RECORD and permission to CLONE. Someone who agreed to be interviewed agreed to be recorded and published as themselves. A synthetic voice that can say things they never said is a different thing, and it needs to be asked for as a different thing.

What makes a good reference recording

A clone reproduces the recording, not the person. That sentence explains every disappointing clone anyone has ever made here: whatever is audible in the sample — the room, the traffic, the second voice, the compression — is treated as part of the voice and comes back in every line it ever reads.

  • One person, speaking alone. Any second voice in the recording — an interviewer, a laugh off-mic — is material the clone will try to reproduce.
  • A quiet room. Background hiss, traffic and air conditioning are part of the voice as far as the clone is concerned, and there is no later setting that removes them.
  • Natural speech, not performance. Read something ordinary at your ordinary pace; a sample recorded in a character voice produces a clone that can only do that character.
  • No music, no processing, no reverb. Anything printed onto the recording is printed onto the clone.
  • Consistent distance from the microphone. A sample where you drift closer and further teaches the clone an inconsistency it will keep.

Tip

Each file must run between 5 and 60 seconds and stay under 10 MB. That is checked in your browser before the upload starts, so a rejected file has not cost you anything — trim it and try again.

Write this
Not this
Where you record
A small carpeted room with the door shut. Soft furnishings beat expensive microphones.
A café, a car, or a room with a fan running.
What you read
Something ordinary at your ordinary pace. The clone learns the voice you gave it.
A dramatic passage performed in a character voice.
How many samples
A few short clips recorded the same way in the same place.
One long clip that drifts between rooms and moods.
Processing
The raw recording. Anything printed onto the file is printed onto the voice.
A podcast export with compression, EQ and a noise gate already applied.

Where you record

Write this:
A small carpeted room with the door shut. Soft furnishings beat expensive microphones.
Not this:
A café, a car, or a room with a fan running.

What you read

Write this:
Something ordinary at your ordinary pace. The clone learns the voice you gave it.
Not this:
A dramatic passage performed in a character voice.

How many samples

Write this:
A few short clips recorded the same way in the same place.
Not this:
One long clip that drifts between rooms and moods.

Processing

Write this:
The raw recording. Anything printed onto the file is printed onto the voice.
Not this:
A podcast export with compression, EQ and a noise gate already applied.

How many clones you can keep

Cloning itself is not behind a paid plan — a free account can make one. What the plan decides is how many you may hold at once.

Plan
Slots
Basic
3 custom voices — The free tier. Cloning is not gated behind a paid plan — the number of clones you may keep is.
Starter
10 custom voices — A legacy plan, no longer sold. Listed because existing subscribers are still on it.
Creator and above
Unlimited — Creator, Pro, Studio and Enterprise carry no cap on custom voices at all.

Custom-voice slots by plan.

Note

The studio shows your own live figure, and the server is what actually enforces it — so if the two ever disagree, the refusal is the truth. Deleting a clone frees its slot straight away, and on 3 slots the usual reason someone is stuck is a test clone made months ago.

When a clone is the wrong tool

A clone is for reproducing one specific person. It is a poor way to shop for a general quality, and it is a genuinely bad way to change how a voice performs — which is what most people are actually trying to do when they reach for it.

If you want
Do this instead
You want a voice that sounds professional
The catalogue already holds 200+ voices recorded and tuned for exactly that. A clone of your own untrained reading will not beat them, and you will spend the slot finding out.
You want a particular accent or age
Filter the catalogue. Cloning is for reproducing ONE specific person; it is a poor way to shop for a general quality, because you have to find a person with that quality first.
You want a different emotion from a voice you like
Direct it. A bracketed direction changes the performance of any catalogue voice, and it costs nothing to try again — where a clone bakes whatever mood you recorded into every line it will ever read.
You want to sound like a famous performer
Do not. It is their voice, it is usually their income, and the fact that a clean recording is easy to find is not a permission — see the consent rule.
You have only a noisy or crowded recording
Record five clean seconds on a phone in a quiet room instead. A short clean sample beats a long dirty one, and the noise in a dirty sample becomes part of the voice permanently.

What to do instead.

Tip

Try directing a catalogue voice before you clone anything. Most of what people hope a clone will fix is a performance problem, and a performance problem is one bracket away — where a clone costs a slot, a recording session, and a result you cannot adjust afterwards.

Once you have one

A cloned voice behaves like any other from here: it takes the same directions, the same pauses and the same phonetic spellings as a catalogue voice, and it is billed the same way. Test it on the text you actually intend to use rather than on the sample script — product names and jargon are where a clone that sounded perfect on prose starts to wobble.

Questions

What people ask about this

Free to start

Now go make one

Reading about a style description only gets you so far. The studio is free to use — write one sentence and hear what comes back.

Create a free account

1,000 credits every month. No card required.

Previous lessonDirecting a performanceNext lessonLong form and multiple speakers

All Voice Studio lessons · Producing a whole book instead

Make something worth hearing.

Start creating free

Create

  • Music
  • Text to speech
  • Audiobooks
  • Podcasts
  • Voice changer
  • Audio reader
  • Narration

Edit & convert

  • Stem splitter
  • Separate speakers
  • Noise reduction
  • Speech to text
  • Media converter
  • Browser DAW
  • All features

Developers

  • Developer hub
  • API reference
  • Quickstart
  • Python SDK
  • MCP server
  • Changelog
  • API status

Resources

  • Guides
  • Languages
  • Use cases
  • Alternatives
  • Tool comparisons
  • AI audio guide
  • Glossary
  • Showcase

Free tools

  • Audio Format Converter
  • Video to Audio Extractor
  • Voice Recorder
  • Free Stem Splitter
  • Free Vocal Remover
  • All free tools

Company

  • About
  • Manifesto
  • Careers
  • Blog
  • Customers
  • Affiliate program
  • Contact
All pages · Sitemap

Studio

  • AI Music & Rap
  • Text to Speech
  • Audiobook Studio
  • Podcast Generator
  • Voice Changer
  • Audio Reader
  • AI Narrator
  • Studio overview
  • All features

Edit & process

  • Stem Splitter
  • Speaker Separation
  • Noise Reduction
  • Speech to Text
  • Media Converter
  • Browser DAW
  • YouTube to Podcast

Voices

  • Voice library
  • Languages
  • Iconic voices
  • Showcase
  • Music Radio

Free tools

  • All free tools
  • Audio Format Converter
  • Video to Audio
  • Audio Trimmer
  • Voice Recorder
  • ACX Checker
  • Free Stem Splitter
  • WAV to MP3 Converter
  • MP4 to MP3 Converter

Solutions

  • Audiobook authors
  • Podcasters
  • Musicians & creators
  • Education
  • Voice agents
  • Gaming
  • Accessibility
  • Advertising
  • All use cases
  • Authors program
  • Enterprise

Compare

  • vs ElevenLabs
  • vs Suno
  • vs Descript
  • vs Murf
  • vs NotebookLM
  • vs LALAL.AI
  • vs NarrationBox
  • All alternatives
  • Tool comparisons

Resources

  • Blog
  • Guides
  • Music Studio guides
  • Audiobook guides
  • Speaker Separation guides
  • Stem Splitter guides
  • Voice Studio guides
  • Transcription guides
  • Changelog
  • Launches
  • Customers
  • Glossary
  • AI Audio guide
  • Family voice (mobile)
  • AudioPod mobile
  • Affiliate program
  • Pricing
  • Developers
  • For AI agents
  • AudioPod for Startups

Company & legal

  • About
  • Manifesto
  • Careers
  • Press & media
  • Contact
  • Responsible AI
  • Voice consent
  • Trust & security
  • System status
  • Security disclosures
  • Security policy
  • Privacy
  • Cookie policy
  • Terms
AudioPod AI

© 2026 AudioPod AI, Inc. All rights reserved.

Privacy|Terms|Trust Center|Responsible AI|Voice consent