Skip to main content
Last verified: 2026-06-11

​
Overview

Genesys partners with Cartesia for text-to-speech in a few ways, one of which is the Genesys TTS Connector integration. This integration is configured in the Genesys Cloud.

​
Prerequisites

​
Configure the Genesys TTS Connector with Sonic

1

Open the Genesys TTS Connector integration

If you did not previously install the integration into your Genesys Cloud organization, install the Genesys TTS Connector now.If you already installed the integration:
  1. Click Menu > IT and Integrations > Integrations.
  2. Search for and open the Genesys TTS Connector integration.
Genesys Cloud integrations page showing the Genesys TTS Connector integration
2

Name the integration

In the Details tab, enter a unique name for the integration. This can be anything you want, such as Genesys TTS Connector: Cartesia.
3

Open the Configuration tab

On the Genesys TTS Connector: Cartesia integration page, click Configuration.
4

Configure properties

Under Properties, enter these values:
Property nameValue
Synthesize Text URIhttps://api.cartesia.ai/tts/bytes
List Voices URIhttps://api.cartesia.ai/voices/export
Synthesize Text Request Methodpost
Voices Attributedata
Voice Name Attributename
Voice Id Attributeid
Voice Language Attributelanguage
Voice Gender Attributegender
PCI CompliantTrue
  • Synthesize Text Attribute
  • Synthesize Voice Attribute
  • Synthesize Languages Attribute
  • Synthesize OutputFormat Attribute
  • Synthesize Sample Rate Attribute
  • Voice Type Attribute
Genesys TTS Connector properties configured for Cartesia
5

Set the advanced JSON configuration

Click Advanced, then copy and paste this as the JSON configuration:
{
  "synthesizeBody": {
    "voice": {
      "mode": "id",
      "id": "$voice"
    },
    "model_id": "sonic-latest",
    "transcript": "$text",
    "output_format": {
      "container": "raw",
      "encoding": "pcm_s16le",
      "sample_rate": 8000
    }
  }
}
  1. If these settings generate garbled audio, come back and change output_format to match your telephony setup. See Output Format for details.
  2. Before using Sonic in production, change sonic-latest to a dated model snapshot from TTS models for stability.
Genesys TTS Connector advanced JSON configuration for Cartesia
6

Add credentials

  1. Click Credentials, then click Configure.
  2. In the Configure Credentials dialog, select Credential Type: User Defined.
  3. Click Add Credential Field, then add these fields:
    KeyX-Api-Key
    X-Api-KeyYour API key from play.cartesia.ai/keys
    KeyCartesia-Version
    Cartesia-Version2026-03-01
  4. Click Confirm.
Genesys TTS Connector credentials configured with Cartesia API headers
7

Save and activate the integration

Click Save & Activate.Back on the Integrations page, your Genesys TTS Connector integration status should change to Active after a few seconds, or show an error.
Error messageWhat it means
You must be logged in to access this endpointX-Api-Key credential is incorrect
Invalid Cartesia-Version headerCartesia-Version credential is incorrect
quota_exceededYou need to upgrade your Cartesia subscription
Your request was invalidSomething is wrong with Configuration > Properties or Configuration > Advanced
Genesys TTS Connector integrations in an error and active state
8

Try out Sonic

On the Genesys Cloud Architect page, select your newly configured integration and use it for a supported language.You can try out different voices on the Architect page, or on play.cartesia.ai/voices. See Choosing a Voice for tips.
If audio is garbled, you may need to change your output format to match your telephony setup:
  1. Read the output format guide
  2. Go back to step 5 and try a different output_format in Configuration > Advanced
Genesys Architect using Cartesia voices through the Genesys TTS Connector integration

​
Resources