Custom voices allows you to use your own voice with Text to Speech, letting you generate clips of your trained voice speaking any text you input. Your ustom voice is found in the Generative Audio "Select a voice" menu, marked with a lightning bolt icon. Click the "Custom" button in the "Select a voice" menu to see all your custom voices. Hover over the name of your custom voice to delete, rename, or share it.

Each custom voice costs 300 credits to train.

Creating a custom voice

1. Click the “Create a custom voice” button in the Generative Audio prompt box, then click “Try it now.”
2. Upload a voice sample between 2 to 5 minutes in length. The sample must be of your voice or someone whose voice you have explicit permission to clone. Ensure your recording has a good range of voice and tone, and reduce background noise as much as possible.
3. If you don't have a pre-recorded voice sample to upload, click "view example” for a script that you can record directly in the app.
4. Click “Name and consent” to proceed, then give your voice a name for easy identification.
5. Click “Submit” to train your custom voice.

Using your cloned voice

Input any text into the Generative Audio prompt box, then click on your custom voice to select it. Custom voices can be used with Lip Sync to give characters distinct and consistent voices.