> [!WARNING]
>
> **Preview**
>
>
> This product or feature is
>
> subject to the "Pre-GA Offerings Terms" in the General Service Terms section
> of the [Service Specific
> Terms](https://docs.cloud.google.com/terms/service-terms#1).
>
> Pre-GA products and features are available "as is" and might have limited support.
>
> For more information, see the
> [launch stage descriptions](https://cloud.google.com/products/#product-launch-stages).

When using text-to-speech for your agent,
you can create personalized voice models by training a model with your own
audio recordings.

## Limitations

This feature is only available for the following languages and locales:

- ar-XA
- cmn-CN
- de-DE
- en-AU
- en-GB
- en-IN
- en-US
- es-ES
- es-US
- fr-CA
- fr-FR
- gu-IN
- hi-IN
- id-ID
- it-IT
- ko-KR
- mr-IN
- nl-NL
- pl-PL
- pt-BR
- ru-RU
- ta-IN
- te-IN
- th-TH
- tr-TR
- ur-IN
- vi-VN

## Configuration

To configure a voice cloning:

1. This feature is provided via the Text-to-Speech product, and the feature is currently restricted access. See the [instructions for requesting access](https://docs.cloud.google.com/text-to-speech/docs/chirp3-instant-custom-voice).
2. [Train your model](https://docs.cloud.google.com/text-to-speech/docs/chirp3-instant-custom-voice#use_chirp_3_instant_custom_voice). When creating your model, copy the voice key. You will need this value to configure your agent.
3. Go to the Dialogflow CX agent settings, select the **Speech and IVR** tab, then scroll down to **Voice selection**.
4. For the voice name, select **Cloned voice**.
5. For the voice key, paste the voice key you copied above.
6. You can click the **Test key** button to verify that the key is valid.