# Clone your voice

Make a copy of your own voice from recordings, then use it for voiceovers. Needs a clear sample and your confirmation that the voice is yours.

Source: https://infinall.ai/docs/audio/voice-cloning
Last checked: 2026-10-26

## What it is

Voice cloning makes a new voice that sounds like a real speaker, from a short recording. You give it a name, add a sample, confirm you have the right to use that voice, and it appears in your voice list, ready for voiceovers.

## Who can use it

Pro only. On Free, **Clone voice** shows a **Pro** badge in the mode list, and choosing it opens an upgrade prompt. The backend also checks your plan before it does anything. Accounts on a trial get the Pro audio features.

## Cost

Cloning a voice is free. Speaking with the cloned voice uses credits, like any premium voice. You see what each take cost on its card. See [Price before you confirm](/docs/credits/price-before-you-confirm).

## Where to find it

On **Home**, choose the **Audio** tab, open the mode chip and pick **Clone voice** ("Train your own custom voice") in the Voice group. It is also in the mode chip of the Audio composer on the Studio canvas.

## Buttons and controls

<!-- shot: audio/voice-cloning -->

| # | Name in the app | What it does | Plan |
| --- | --- | --- | --- |
| 1 | Clone voice | The mode chip, after you pick Clone voice. | Pro |
| 2 | Name this voice, e.g. My narrator voice | The name box. Give the voice a name you will recognise. | Pro |
| 3 | Upload files | A tab. Add finished recordings from your computer. | Pro |
| 4 | Record | A tab. Record in the browser, reading a short script aloud. | Pro |
| 5 | Upload recordings of the voice | The upload area on the **Upload files** tab. Click it or drop files on it. | Pro |
| 6 | This is my voice, or I have the speaker's permission to clone it. | The consent tick box. You must tick it. | Pro |
| 7 | Clone voice | The big button at the bottom, which says **free** after the name. It makes the voice. While it works it says **Cloning...**. It stays off until the name, a sample and the consent tick are all in place. | Pro |
| | Remove (file name) | The X next to each uploaded file. Takes that file out of the list. | Pro |
| | Read this aloud | On the **Record** tab. A short script to read while you record. | Pro |
| | Start recording, Stop, Record another take | On the **Record** tab. Starts, stops and adds takes. | Pro |
| | Delete take 1 | A bin icon next to each recorded take. Takes that recording out. | Pro |
| | Your voices | A list of the voices you made, each with **Ready** or **Failed**. Only shows once you have one. | Pro |
| | Delete (voice name) | The bin icon beside a voice in **Your voices**. Deletes the voice. | Pro |

## How to clone your voice from files

1. Open **Home**, choose **Audio**, then pick **Clone voice** in the mode chip.
2. Type a name in **Name this voice**.
3. Stay on **Upload files**. Click **Upload recordings of the voice** and pick your recordings, or drop them on it.
4. Tick **This is my voice, or I have the speaker's permission to clone it.**
5. Choose the big **Clone voice** button at the bottom (it says free after the name).
6. When it is done, the voice shows **Ready** in **Your voices**, and the composer switches to **Voice over** with your new voice already picked.

## How to clone your voice by recording

1. Pick **Clone voice**, then type a name.
2. Choose the **Record** tab.
3. Choose **Start recording** and read the script under **Read this aloud**, in a quiet room at your normal pace.
4. Choose **Stop**. Add more takes with **Record another take** if you need more speech.
5. Listen back to each take. Remove any bad one with its bin icon.
6. Tick the consent box and choose the big **Clone voice** button.

The hint under the recorder counts your time. It tells you how many seconds you have, whether you can clone yet, and when you have "plenty".

## How to speak with your cloned voice

1. Open the mode chip and choose **Voice over**.
2. Open **Voice**. Your clones are listed under **Your voices**, with the tone "Cloned voice".
3. Pick your voice, type a script and send it. The speech model is picked for you.

Emotion tags and the emotion chip do not work on a cloned voice. Everything else in [Make a voiceover from a script](/docs/audio/make-a-voiceover) does.

## How to delete a cloned voice

1. Open **Clone voice**.
2. In **Your voices**, choose the bin icon (**Delete** followed by the voice name) next to the voice.

## Options explained

**Upload or record.** Both ways make the same kind of request. Use whichever is easier. Recording asks for your microphone, and its script has varied sounds and sentences, which helps the clone.

**Audio files.** MP3, WAV, M4A and OGG files work. You can add up to ten files.

**What makes a good sample.** Clear speech with little background noise, in your normal voice. A minute or two sounds best. When you record, 30 seconds or more sounds better than the minimum.

**Consent.** You must confirm that this is your voice or that you have the speaker's permission before cloning. Without the tick, the big **Clone voice** button stays off.

**Who can see it.** A cloned voice belongs to the workspace that made it. Your teammates in that workspace can use it, and it can be deleted from the list.

## Limits

- A clone needs 1 to 10 recordings, with at least 10 seconds of clear speech in total.
- Each file can be up to 50 MB.
- When recording: up to 3 takes, each up to 120 seconds, with at least 10 seconds in total. 30 seconds or more sounds better.
- A voice name can be up to 128 characters.
- Your account has a limited number of voice slots. If they are all used, delete a clone to free one.

## When something goes wrong

| Message in the app | What it means | What to do |
| --- | --- | --- |
| Premium voices, music, sound effects, voice changer, noise removal, voice cloning and dubbing are Pro features. Upgrade to Pro to use them. | You tried to clone on a Free account. | Upgrade to Pro. |
| Upload at least one recording of the voice. | No file was added. | Add a recording, or record one. |
| `That's <n> seconds of audio. Upload at least 10 seconds of clear speech. A minute or more sounds best.` | Your samples add up to under 10 seconds. | Add more speech. |
| That sample is too short. Upload at least 10 seconds of speech. | The voice service found too little speech in the sample. | Add a longer, clearer recording. |
| All voice slots are in use. Delete a cloned voice and try again. | You have used every voice slot. | Delete a voice you do not need in **Your voices**, then try again. |
| Voice cloning failed. | The clone did not finish. | Try again with a clearer sample. The voice shows **Failed** in **Your voices**, and you can delete it. |
| Couldn't clone that voice. | The request failed. | Try again. |
| `<file name> isn't an audio file. Upload MP3, WAV, M4A or OGG.` | The file is not an audio file. | Pick an audio file. |
| `<file name> is over 50 MB.` | The file is too big. | Use a shorter or smaller recording. |
| This browser can't record audio. Upload a recording instead. | Your browser cannot use a microphone for recording. | Use the **Upload files** tab. |
| Microphone access is blocked. Allow the microphone for this site in your browser, then try again. | The browser blocked the microphone. | Allow the microphone for infinall.ai in your browser settings. |
| No microphone found. Plug one in, or upload a recording instead. | No microphone was found. | Plug one in, or upload a file. |
| The microphone is in use by another app. Close it and try again. | Another app has the microphone. | Close the other app. |
| Couldn't delete that voice. | The voice could not be deleted. | Try again. |
| Voice not found. | The voice was already deleted. | Refresh the list. |

## Questions

### Is it okay to clone someone else's voice?

Only with their permission. You must tick the box confirming that the voice is yours or that you have the speaker's permission. Without it, cloning does not start.

### Does cloning cost credits?

No, it is free. Speaking with the cloned voice uses credits, like any premium voice.

### How long should the sample be?

At least 10 seconds of clear speech in total. One to two minutes sounds best.

### Why did my clone say Failed?

The sample may have been too short or too noisy. Delete the voice, then try again with a longer, cleaner recording.

### Can my team use my cloned voice?

A cloned voice belongs to the workspace it was made in, so it shows up in the voice list for that workspace.

### Can I use emotions with a cloned voice?

No. Emotions work only on voices tagged **EMOTIONS**.
