Use more than one voice

Assign multiple voices to simulate a team.

Use multi-voice when you want callers to hear different voices across interactions — simulating a team of agents or varying the experience for repeat callers. You can configure this through the UI or programmatically in Python.

Add voices

Click New voice to add additional voices to your agent. You can select from the available languages, genders, and styles or add a new voice.

Once added, all voices will appear in the list and can be configured independently.

Adjust weights

Adjust voice weights

After adding multiple voices, use the weight sliders to define how frequently each voice is selected during conversations. All weights must add up to 100%.

For example:

• 80% main voice

• 10% each for two alternates

  

Configure settings

Customize individual voice settings

For each voice, you can fine-tune the advanced settings:

• Stability (%): Control how consistent the voice sounds across generations.

• Clarity and Similarity (%): Balance naturalness and similarity to the target voice.

  

Using Python in a code field

You can configure this in any function, though this guide assumes you’re using the start function.

To use multiple voices, you need their voice IDs from your chosen TTS provider. Refer to the function TTS provider configuration for full details on supported providers and their configuration options.

Here is an example of how to configure multiple voices, where the agent is randomly assigned a voice at the start of each interaction:

def start_function(conv: Conversation):
    conv.randomize_voice([
        VoiceWeighting(
            voice=ElevenLabsVoice(
                voice_id="LcfCDJNUPlGQjkzn1xUU",
                similarity_boost=0.2,
                stability=0.4
            ),
            weight=0.5
        ),
        VoiceWeighting(
            voice=RimeVoice(
                provider_voice_id="s3://path-to-manifest.json",
                style="neutral"
            ),
            weight=0.5
        )
    ])

• conv.randomize_voice([...]): Selects a voice at random based on the assigned weights.
• VoiceWeighting: Associates a voice with a probability of being selected.
• Weights define selection probability:
  • The sum of all weights must equal 1.0.

Understanding voice weights

You can add a maximum of ten voices to an AI agent.

In the UI, weights are displayed as percentages (must total 100%). In code, weights are decimal values (must total 1.0).

The weight parameter determines how often each voice is selected. The sum of all weights must equal 1.0 (or 100% in the UI).

Examples:

• One voice at 1.0: Always uses that voice.
• Two voices at 0.5 each: Each is selected 50% of the time.
• Four voices at 0.25 each.

Personalizing repeat caller experiences

Assigning different voices to repeat callers can help create a sense of interacting with different team members, which may reduce escalation requests. Use the conv.randomize_voice() method in your start function to vary voice selection across interactions.