Last updated
What is Deepgram Aura used for in a Bolna voice agent?
Deepgram Aura is Deepgram's text-to-speech model family, and in Bolna it turns your agent's English replies into speech with the aura-2 model. Bolna's provider guide lists it as fast and accurate for English.
Pick it when your agent speaks English and you already use Deepgram for transcription.
How do I set up Deepgram Aura with Bolna?
- Optional: connect your own Deepgram account at platform.bolna.ai/auth/deepgram.
- Open your agent in Agent Studio and go to the Languages tab, or build the agent through the Create Agent API.
- Set the synthesizer provider to
deepgramand the model toaura-2. - Pick a voice. Resolve valid voices with the List voices endpoint if you use the API.
- Leave
languageout ofprovider_config. Deepgram does not take it. - Turn on
stream: trueand place a test call.
For more detail, see the Deepgram Aura setup guide in the Bolna docs.
Which Deepgram TTS models does Bolna support?
| Model | Use | Notes |
|---|---|---|
aura-2 | English conversational speech | Listed in the Create Agent API reference |
aura | Earlier Aura model | Listed on the Deepgram voice page only |
Use aura-2 for new agents, since it is the model the Create Agent API reference documents.
What can you build with Deepgram Aura and Bolna?
An all-Deepgram English agent. Use Deepgram Nova-3 or Flux for transcription and Aura for the voice, so one provider account covers both sides of the call.
English support and sales lines. Aura suits English conversational calls such as inbound support, appointment booking with Cal.com, or lead qualification.
An English voice alongside Indian languages. In a multilingual agent, run Aura on the English tab and Sarvam on Hindi or other Indian-language tabs.
Comparing voices on the same agent. Try Aura against Cartesia or ElevenLabs on test calls and keep the one that suits your callers.