For the agent created, you can configure its language settings. Currently that setting would have the following effects:

  • Determine what language the agent would understand (language of speech to text)
  • Determine the default TTS model choice based on that language (default voice model)
  • Will pass that language code to TTS provider and influence the synthesized speech. For multilingual, it will auto detect the language based on the generated text.
  • Influence the speech normalization process (see Normalize text for speech)

What it does not do:

  • It will not enforce the agent to generate responses in a certain language. You have to set that in the prompt yourself. This can be as easy as adding something like always respond in Spanish into the prompt.
  • It will not enforce the agent to pronounce the generated text in a certain accent. You have to find a voice that speaks that accent.

​
Language supported

Check the dashboard for most up to date list of language support. Every quarter we are adding support for more languages.

For the multilingual option, it will be able to understand and codeswitch between a variety of languages. Check the list of languages supported for that option by hovering over the option.

​
Set language programmatically

You can set the language of the agent in the Create Agent API or Update Agent API by setting the language field to the desired language code.