Custom Voices
for Enterprises

To train a Custom Voice, you will need 60-90 minutes of clear and high-quality voice recordings. However, for many Target Voices, good results can be obtained from only 30 minutes of audio. The amount of training data required depends on the proximity of the Target Voice to Altered's dataset. Altered also offers special custom services with less that 30 min of audio, but at a higher cost.

Rapid Voice Cloning is a quick process of creating a voice model in minutes, typically used for fast prototyping or testing. Custom Voice, on the other hand, involves creating a more refined and tailored AI-generated voice model for a specific brand or project, ensuring excellent vocal likeness. Rapid Voice Cloning requires only a few seconds of training data and is available at all times to all users at no cost, but has lower likeness than a full Custom Voice.

A Custom Voice is an AI-generated voice model tailored to a specific brand, project, or individual, providing a unique and consistent vocal identity for various applications like advertising, content creation, and more. A Custom Voice is trained by Altered for you and requires a larger amount of high quality data than Rapid Voice Cloning. Custom Voice models can be trained in both Speech-To-Speech and Text-To-Speech.

Yes, you can use audio extracted from a video to clone your voice with Altered Studio. In fact, you can use Altered Studio to extract that audio with ease.

Altered Studio Professional and Enterprise plans allow unlimited number of Custom Voices. Altered Studio Creator plan has a limit on the number of Custom Voices you can have in your library at a time.