Skip to main content
Enterprise customers can commission a custom voice built from recorded samples of a specific speaker, giving your agent a unique, brand-specific sound that only your organization uses.

​
What custom voices offer

  • Brand-exclusive voice – A synthetic voice trained on recordings of a specific speaker, available only to your organization.
  • Consistent identity – The same voice across all calls, environments, and languages (where supported).
  • Fine-tuned characteristics – Control over tone, pace, and delivery style during the cloning process.

​
Requirements

  • An active PolyAI enterprise account.
  • High-quality audio recordings of the target speaker (PolyAI provides recording specifications).
  • Legal clearance from the speaker to use their voice for synthesis.

​
How to request

  1. Contact your PolyAI account manager or representative.
  2. PolyAI provides a recording brief with duration, format, and content requirements.
  3. After recordings are submitted, the voice is trained and made available in your Voice Library.
Custom voice creation typically takes 2-4 weeks depending on recording availability and review cycles.

Voice library

Browse available voices

Agent Voice

Configure voice selection and settings

Choosing a good voice

Guidelines for voice selection

Voice overview

Complete voice configuration guide
Last modified on April 16, 2026