> ## Documentation Index
> Fetch the complete documentation index at: https://docs.percify.io/llms.txt
> Use this file to discover all available pages before exploring further.
# Clone your voice with AI
> Record or upload a 5 to 60 second sample of your voice in Percify Voice Studio, type a script, and generate speech in your cloned voice for 5 credits.
Percify Voice Studio clones your voice from a short sample. Record 5 to 60 seconds of yourself talking (or upload an audio file), type what you want said, and Percify generates that speech in your voice. Each generated clip costs 5 credits, and the clips are saved so you can use them in talking avatar videos.
## What you need
* A computer. Voice Studio runs in the desktop web app at [app.percify.io/home/voices](https://app.percify.io/home/voices), from the **Voice Studio** tile on the Explore page.
* A microphone, or an audio file of your voice under 10 MB.
* The script you want read, in English.
* At least 5 credits.
## How to clone your voice
On Voice Studio, click **Create or Clone a Voice**.
In the right-hand column, click **Tap to record** and read something naturally for 5 to 60 seconds (recording starts after a 3 second countdown and stops at 60 seconds). Or click **Upload** and pick an audio file. When the sample is ready you see "Voice clip ready", and the footer reads "Generating with your cloned voice".
Write what your voice should say in the text box.
Click **Generate**. When the clip is ready, play it in the preview. The **History** tab lists every clip from this session with a download button.
The sample stays active only while the window is open. Next time you open **Create or Clone a Voice**, record or upload it again.
## How much does it cost?
| What | Credits |
| ---------------------------------------------- | --------- |
| Uploading or recording a sample | No charge |
| Each generated clip in your cloned voice | 5 |
| Each generated clip with a Voice Library voice | 5 |
You are charged after the clip is generated. If generation fails, nothing is charged. See [how credits work](/percify/credits).
## Limits
| Limit | Value |
| --------------- | ---------------------------------------------------------------------------- |
| Recorded sample | 5 to 60 seconds |
| Uploaded sample | An audio file under 10 MB |
| Language | Cloned speech is generated in English. Voice Studio has no language setting. |
## Where do my clips go?
Every generated clip is saved to your account. You find it under **My Voices** in Voice Studio, and under **My Voices** when you make a talking video, so you can [put your cloned voice on a photo](/create/clone-yourself).
## Tips for a good clone
* Record in a quiet room, close to the microphone, with no music behind you.
* Speak the way you want the clone to sound: same pace, same energy.
* Use most of the 60 seconds rather than the 5 second minimum.
* If recording fails, check that the browser is allowed to use the microphone.
Not from Voice Studio. The cloned speech is generated in English and there is no language option. To translate a finished video, see [Dub a video into another language](/create/dubbing).
Your sample is uploaded to your account, and every clip you generate is saved. The clone itself is not kept as a named voice: when you reopen the workspace, add the sample again.
No. Podcast Studio uses its own 10 preset voices. Use your clips in the talking avatar generator instead.
## Related
Put your cloned voice on a photo of yourself.
Turn any recording into a talking video.
Fixes when the clone sounds wrong or will not record.
Plans, monthly credits and refunds.