AI Assistants Overview

Voice Selection & Voice Cloning

Voice Selection & Voice Cloning

Learn how to select built-in voices or clone your own voice for your AI assistants

Larsa AI – Voice Options for Your AI Assistant

Your AI assistant can speak using either built-in voices or a custom cloned voice.
Natural, realistic voices help increase customer trust and engagement.


1. Built-In Voice Library

  • Powered by: ElevenLabs.

  • Selection: Choose male or female voices, accents, and languages from the default list.

  • Customization: Adjust pitch, speed, or filler usage if needed.



2. Cloning Your Own Voice

Why Clone?
  • Maintain brand consistency.

  • Use a familiar voice that resonates with your audience.


Audio Sample Requirements:

Clear recording, usually 1–5 minutes in length.

Process:
  1. Record yourself reading a provided script (no background noise).

  2. Upload the recording in the Clone Voice section.

  3. Wait for the training to complete (could take minutes or hours).

  4. Test the new voice with short sample texts.



3. Best Practices

  • High-Quality Audio: The cleaner your sample, the better the cloned result.

  • Steady Delivery: Speak in a natural, consistent tone without abrupt changes.

  • Legal Check: Make sure you have the rights or permission to clone any voice that isn’t your own.

Custom Mid-Call Tools

Cal.com Appointment Scheduling

© 2025 Larsa AI LTD. All rights reserved.