Multi-Voice SDK

A universal Text-to-Speech (TTS) and Speech-to-Text (STT) SDK that supports multiple providers including Google Gemini, Deepgram, OpenAI, Groq PlayAI, Cartesia, and AssemblyAI. Easily generate audio content, transcribe speech, and manage audio files with a unified API.

Features

• 🎵 Multi-Provider TTS: Gemini, Deepgram, OpenAI, Groq PlayAI, and Cartesia TTS
• 🎙️ Speech-to-Text: Deepgram and AssemblyAI STT with advanced features
• 🔧 Audio Merging: Combine multiple audio files seamlessly
• 🎯 Simple API: Easy-to-use functions with consistent interface
• 📦 ESM Ready: Modern ES modules support

Installation

npm install multi-voice-sdk

Quick Start

import { tts, stt, merge } from "multi-voice-sdk";

// Generate speech with OpenAI
tts({
  provider: "openai",
  apiKey: "your-api-key",
  text: "Hello, world!",
  voice: "nova",
  outputFile: "output.mp3",
});

// Transcribe audio with Deepgram
stt({
  apiKey: "your-deepgram-key",
  audioFile: "https://example.com/audio.wav", // Can be URL or local file
});

// Merge multiple audio files
merge({
  inputFiles: ["file1.mp3", "file2.mp3"],
  outputFile: "combined.mp3",
});

API Reference

tts(options)

Generate speech from text using various TTS providers.

Parameters

| Parameter | Type | Required | Description | | ------------ | -------- | -------- | ----------------------------------------------------------------------------- | | provider | string | ✅ | TTS provider: "gemini", "deepgram", "openai", "groq", or "cartesia" | | apiKey | string | ✅ | API key for the chosen provider | | text | string | ✅ | Text to convert to speech | | voice | string | ✅ | Voice identifier (provider-specific, for Cartesia use voice ID) | | outputFile | string | optional | Output file path (default: "output.mp3") | | model | string | optional | Model to use (provider-specific) | | prompt | string | optional | Additional instructions for speech generation |

Examples

OpenAI TTS

tts({
  provider: "openai",
  apiKey: process.env.OPENAI_API_KEY,
  model: "gpt-4o-mini-tts",
  text: "Hello from OpenAI!",
  voice: "nova",
  prompt: "Speak in a cheerful tone",
  outputFile: "openai_output.mp3",
});

Google Gemini TTS

tts({
  provider: "gemini",
  apiKey: process.env.GEMINI_API_KEY,
  text: "Hello from Gemini!",
  voice: "iapetus",
  prompt: "In a pleasant and calm tone",
  outputFile: "gemini_output.mp3",
});

Deepgram TTS

tts({
  provider: "deepgram",
  apiKey: process.env.DEEPGRAM_API_KEY,
  text: "Hello from Deepgram!",
  voice: "aura-2-luna-en",
  outputFile: "deepgram_output.mp3",
});

Groq PlayAI TTS

tts({
  provider: "groq",
  apiKey: process.env.GROQ_API_KEY,
  text: "Hello from Groq PlayAI!",
  voice: "Arista-PlayAI",
  outputFile: "groq_output.wav",
});

Cartesia TTS

tts({
  provider: "cartesia",
  apiKey: process.env.CARTESIA_API_KEY,
  text: "Hello from Cartesia!",
  voice: "694f9389-aac1-45b6-b726-9d9369183238", // Voice ID
  outputFile: "cartesia_output.mp3",
});

stt(options)

Transcribe audio to text using Speech-to-Text providers.

Parameters

| Parameter | Type | Required | Description | | ----------------- | --------- | -------- | ------------------------------------------------------------------------- | | provider | string | ✅ | STT provider: "deepgram" or "assemblyai" | | apiKey | string | ✅ | API key for the chosen provider | | audioFile | string | ✅ | Path to local audio file or URL of remote audio file to transcribe | | outputFile | string | optional | Output file path for results (default: "transcription.json") | | model | string | optional | Model to use (default: "nova-3") | | smartFormat | boolean | optional | Enable smart formatting (default: true) | | detect_language | boolean | optional | Automatic language detection (default: true) | | punctuate | boolean | optional | Enable punctuation (default: true) | | diarize | boolean | optional | Enable speaker diarization (default: false) | | channels | number | optional | Number of audio channels (default: 1) | | fullResponse | boolean | optional | Return full response object instead of just transcript (default: false) |

Returns

• Default: Returns transcript as a string
• With fullResponse: true: Returns object with transcript, confidence, words, and metadata

Examples

Deepgram : Basic Transcription (Remote URL)

stt({
  provider: "deepgram",
  apiKey: process.env.DEEPGRAM_API_KEY,
  audioFile: "https://example.com/audio.wav", // Remote URL
});

Deepgram : Local File Transcription

stt({
  provider: "deepgram",
  apiKey: process.env.DEEPGRAM_API_KEY,
  audioFile: "./my-audio.mp3", // Local file path
  outputFile: "transcription.json",
});

AssemblyAI : Basic Transcription (Remote URL)

stt({
  provider: "assemblyai",
  apiKey: process.env.ASSEMBLYAI_API_KEY,
  audioFile: "https://example.com/audio.wav", // Remote URL
  outputFile: "transcription.json",
});

AssemblyAI : Local File Transcription

stt({
  provider: "assemblyai",
  apiKey: process.env.ASSEMBLYAI_API_KEY,
  audioFile: "./my-audio.mp3", // Local file path
  outputFile: "transcription.json",
  fullResponse: true, // Get detailed response
});

merge(options)

Merge multiple audio files into a single file.

Parameters

| Parameter | Type | Required | Description | | ------------ | ---------- | -------- | ------------------------- | | inputFiles | string[] | ✅ | Array of input file paths | | outputFile | string | ✅ | Output file path |

Example

merge({
  inputFiles: ["intro.mp3", "main.mp3", "outro.mp3"],
  outputFile: "complete_audio.mp3",
});

Supported Voices

OpenAI

• alloy, ash, ballad, coral, echo, fable, onyx, nova, sage, shimmer, verse

Gemini

• zephyr (Bright), puck (Upbeat), charon (Informative), kore (Firm), fenrir (Excitable), leda (Youthful), orus (Firm), aoede (Breezy), autonoe (Bright), enceladus (Breathy), iapetus (Clear)

For a complete list of available Gemini voices, see: Gemini Speech Generation Documentation

Deepgram

• aura-2-luna-en, aura-2-stella-en, aura-2-arcas-en, and more

For a complete list of available Deepgram voices, see: Deepgram TTS Models Documentation

Groq PlayAI

• Atlas-PlayAI, Arista-PlayAI, Basil-PlayAI, Briggs-PlayAI, and more

For a complete list of available Groq PlayAI voices, see: Groq TTS Documentation

Cartesia

Cartesia uses voice IDs instead of voice names. Example voice IDs:

• 694f9389-aac1-45b6-b726-9d9369183238 (Default voice)
• Use the Cartesia console to find available voice IDs for your account

For more information about Cartesia voices, see: Cartesia Console

Environment Variables

Create a .env file in your project root:

OPENAI_API_KEY=your_openai_api_key
GEMINI_API_KEY=your_gemini_api_key
DEEPGRAM_API_KEY=your_deepgram_api_key
GROQ_API_KEY=your_groq_api_key
CARTESIA_API_KEY=your_cartesia_api_key

Requirements

• Node.js 16.x or higher

License

ISC