Skip to main content

Open the Voices page

The Voices section is where you manage how your agents sound. From here, you can:
  • Browse the Voice Library (ready-to-use voices)
  • Preview voices before using them
  • Assign a voice to a specific agent
  • Clone a voice (so your agent can speak with a custom voice)
Go to the Voices section by clicking here: Voices
The Voices page showing the Try Voices and Clone Voice tabs, the search bar, the gender and language filters, and the Voice Library list with an Assign button on each voice.

Screenshot: The Voice Library in the Try Voices tab, with the search bar and the gender and language filters.

Browse and preview voices (Try Voices)

By default, you’ll land on Try Voices. This is the Voice Library: a collection of professional AI voices (more than 150), split into pages. Each row includes:
  • A Play button to preview the voice
  • The voice name plus a short style description (for example: “Laid-Back”, “Reassuring”, “Energetic”)
  • Details tags: gender, language and accent (for example: Male, en, american)
  • An Assign button on the right to use that voice with an agent
Use the controls above the list to narrow it down:
  • Search voices by name or accent: type a name (for example “Sarah”) or an accent (for example “british”).
  • Gender: All Genders, Male, Female or Other.
  • Language: pick a language from the list. You can search it by name or by code (for example “Spanish” or “es”). Choose All to see every language again.
When previewing voices, use the same short script each time so it’s easy to compare.Example: “Hi! This is Giulia from Clinica Aurora. I’m calling to confirm your appointment for Thursday at 3 PM.”

Assign a voice to an agent

To make an agent speak with a specific voice:
1
In Try Voices, find the voice you want. Use the filters to get there faster.
2
Click Assign on the right side of that voice.
3
The Assign Voice to Agent window opens. At the top, you’ll see the voice card: its name, a full description and tags such as gender, language, accent, age and use case.
4
Under Select an agent to assign this voice, find the agent you want to update and click the Assign button next to it.

Video: filter the library by gender and language, then click Assign and choose the agent.

When you assign a voice, this window opens with the list of your agents:
The Assign Voice to Agent window with the voice card at the top and the list of agents, each with its own Assign button.

Screenshot: The Assign Voice to Agent window. Each agent in the list has its own Assign button.

After you assign it, that agent will use the selected voice when speaking on calls.
Voices are set per agent (not global). To change an agent’s voice later, assign a different one from here, or open the agent and change it in the Agent tab → Voice. See AI Agents.

Clone a voice (Clone Voice)

If you want a more custom sound (for a brand voice or a specific persona), use Clone Voice.
1
Open the Clone Voice tab.
2
Click Clone Voice.
3
In the Voice Preview window, fill in:
  • Voice Name (required)
  • Description (optional)
  • Language (required): English, Italian, French, German, Portuguese, Spanish, Dutch, Hindi or Arabic
  • Gender (required): Male, Female or Other
4
Add a Voice Sample (required) using one of the options shown:
  • Upload Audio: click Select File or drag and drop an audio file (max. 4MB)
  • Record Voice: record the sample directly in your browser
5
Click Clone Voice to submit your sample. Once your cloned voice is created, it will appear in the Clone Voice tab and can be assigned like any other voice.

Video: open Clone Voice, fill in the name, description, language and gender, then add the voice sample.

Once the form is filled in, the preview window is ready for your voice sample:
The Voice Preview window with the Voice Name, Description, Language and Gender fields, the Upload Audio and Record Voice options, and the Voice Sample upload area.

Screenshot: The Voice Preview window, ready for the voice sample.

Common mistake: Uploading a clip that’s too short or has background noise. Your clone will sound noticeably worse if the sample isn’t clean.

Voice cloning guidelines and best practices

When recording or uploading your sample, follow these guidelines (they also appear at the bottom of the Clone Voice tab):
  • Record in a quiet environment with minimal background noise
  • Speak clearly at a natural pace
  • For uploaded files, provide at least 30 seconds of continuous speech
  • Maximum file size is 4MB
  • Supported languages: English, Italian, French, German, Portuguese, Spanish, Dutch, Hindi and Arabic
Only upload audio you have permission to use (for example, your own voice or a voice you’re licensed to use).