Day 0 support for Google Gemini 3.1 Flash TTS Try it now →
Providers

Deepgram

Deepgram Aura text-to-speech.

Prefixdeepgram
Default modelaura-2
Env varDEEPGRAM_API_KEY
Official docsdevelopers.deepgram.com/docs/text-to-speech

Models

ModelStreamingAudio TagsVoice CloningNotes
aura-2YesNoNoCurrent default
auraYesNoNoPrevious generation

Usage

import { generateSpeech } from "@speech-sdk/core"

const result = await generateSpeech({
  model: "deepgram/aura-2",
  text: "Hello from SpeechSDK!",
  voice: "thalia-en",
})

Voice + Model

Deepgram's API expects the voice to be concatenated with the model (e.g. aura-2-thalia-en). SpeechSDK builds this identifier for you from the voice parameter — just pass the voice suffix like thalia-en, apollo-en, or orion-en.

Provider Options

await generateSpeech({
  model: "deepgram/aura-2",
  text: "Hello!",
  voice: "thalia-en",
  providerOptions: {
    encoding: "mp3",
    sample_rate: 24_000,
    container: "none",
  },
})

Custom Configuration

import { generateSpeech } from "@speech-sdk/core"
import { createDeepgram } from "@speech-sdk/core/providers"

const deepgram = createDeepgram({
  apiKey: process.env.DEEPGRAM_API_KEY,
})

const result = await generateSpeech({
  model: deepgram("aura-2"),
  text: "Hello!",
  voice: "thalia-en",
})

On this page