Cartesia

Example Usage

D-ID provides Cartesia integration to generate text to speech for agents using V4 Expressive Avatars.

{
  "presenter": {
    "type": "expressive",
    "presenter_id": "YOUR_EXPRESSIVE_AVATAR_ID",
    "voice": {
      "type": "cartesia",
      "voice_id": "694f9389-aac1-45b6-b726-9d9369183238"
    }
  }
}

{
  "presenter": {
    "type": "expressive",
    "presenter_id": "YOUR_EXPRESSIVE_AVATAR_ID",
    "voice": {
      "type": "cartesia",
      "voice_id": "694f9389-aac1-45b6-b726-9d9369183238",
      "model_id": "sonic-3",
      "language": "en",
      "voice_config": {
        "rate": "1.2"
      }
    }
  }
}

Available Voices

Use the /voices endpoint with provider=cartesia to get all the supported Cartesia voices.

Voice Config

Speed: The rate parameter controls how fast the voice speaks. The value is relative to 1 (normal speed), for example "0.8" is slower and "1.2" is faster. Valid range is 0.6 to 1.5. Default value: 1. Speed control is supported by the Sonic 3 models and newer (sonic-3, sonic-3.5, sonic-3.6), not by sonic-2.

Model: The model_id parameter selects the Cartesia model. When omitted, the default Cartesia model is used.

Dated snapshot ids such as sonic-3.6-2026-08-27 and the sonic-latest alias are accepted as well. Please visit the Cartesia models page for more information on all the available models

Language: The language parameter is the language of the selected voice in ISO 639-1 (2-letter) format, for example en.