​
Introduction

Voices determine how your AI agent sounds on calls. From natural-sounding English voices to expressive multilingual options, selecting the right voice can dramatically improve the user experience.

With our voice settings, you can:

  • Choose from a wide range of tones, accents, and cadences.
  • Select voices optimized for English or other languages.
  • Enable real-time language switching for multilingual callers.
  • Pair voices with specific AI models for best performance.

Voice configuration is quick to set up and can be customized per agent to match your brand’s tone and target audience.

​
Assigning a Voice in Your Workflow

To assign a voice to your AI agent:

  1. Click on the inbound or outbound agent node.
  2. Select Configure.
  3. Scroll down to the Voice dropdown menu.
  4. Choose your desired voice from the list.

​
Selecting the Right Voice

We offer a wide selection of voices across various languages, tones, speech patterns, and cadences.

  • For English, the highest-quality voices are tagged with EN
  • For non-English languages, we recommend voices tagged with Turbo Multi for the best performance. Enabling Multilingual Capabilities alongside this will ensure the best possible multilingual experience.

​
Multilingual Capabilities

The platform supports multi-lingual conversations through live-switching, which enables the bot to automatically detect the caller’s spoken language and respond in the same language. This allows for seamless interactions without requiring users to select a language upfront.

Live-switching can be enabled by navigating to your Agent Node → Advanced → Multi-lingual —> Toggle Yes

Then, in your prompt node add something like:

Prompt: If the caller speaks in Spanish, respond only in Spanish.

Currently supported languages:

  • English
  • Spanish
  • French
  • German
  • Hindi
  • Russian
  • Portuguese
  • Japanese
  • Italian
  • Dutch

​
Test & Preview Voices

To try out different voices before assigning them:

  1. Navigate to Organization Settings (top-left corner of the dashboard).
  1. Click on Voices in the sidebar.
  2. Use the Voice Playground to preview voices with various example excerpts.

Note: Voices may sound slightly different in the playground compared to production. For the most accurate preview, test voices directly within a live workflow.

​
Need Help?

Join the HappyRobot Slack Community