TTS Models — Deepgram | Documentation

Voice Agent

By default Deepgram Text-to-Speech will be used with the Voice Agent API, but if you opt to use another provider’s TTS model with your Agent, you can do so by applying the following settings.

Deepgram TTS models

These are just some of our most popular TTS voices. For a complete list see the TTS Models Documentation .

agent.speak.model	Gender
`aura-stella-en`	Female
`aura-asteria-en`	Female
`aura-luna-en`	Female
`aura-arcas-en`	Male

Example

JSON

1 {
2   "speak": {
3     "model": "aura-asteria-en"
4   }
5 }

You can set your Text-to-Speech model in the Setting Configuration for your Voice Agent. See the docs for more information.

Third Party TTS models

To use a third party TTS voice, specify the TTS provider and one of their available voice IDs.

For ElevenLabs you can refer to this article on how to find your Voice ID or use their API to retrieve it. We support any of ElevenLabs’ Turbo 2.5 voices to ensure low latency interactions.

For Cartesia you can use their API to retrieve a voice ID.

agent.speak.provider	speak.voice_id
`eleven_labs`	`unique_voice_id`
`cartesia`	`unique_voice_id`

Example

JSON

1 {
2   "speak": {
3     "provider": "eleven_labs",
4     "voice_id": "bIHbv24MWmeRgasZH58o"
5   }
6 }

What’s Next

Configure the Voice Agent

1	{
2	"speak": {
3	"provider": "eleven_labs",
4	"voice_id": "bIHbv24MWmeRgasZH58o"
5	}
6	}