MusicGPT SDK

Official TypeScript package for the MusicGPT API - AI-powered music generation, voice conversion, and audio processing.

MusicGPT public Api

https://musicgpt.com/api

📚 Documentation

https://docs.musicgpt.com/

🎵 Features

• AI Music Generation - Create original music from text prompts
• Voice Conversion - Transform voices with AI technology
• Audio Processing - Denoise, enhance, and modify audio files
• Voice Cloning - Create covers with different voices
• Audio Transcription - Convert speech to text
• Music Analysis - Extract key, BPM, and stems from audio
• TypeScript Support - Full type safety and IntelliSense
• Async/Await - Modern Promise-based API

📦 Installation

npm install musicgpt
# or
yarn add musicgpt

🚀 Quick Start

import { MusicGPTClient, MusicGPTConversionType } from 'musicgpt';

const client = new MusicGPTClient('your-api-key');

// Generate music from text
const music = await client.music_ai({
  prompt: 'A cheerful pop song about summer',
  music_style: 'pop',
  make_instrumental: false,
});

// Wait for completion
const result = await client.waitForCompletion(music.task_id, MusicGPTConversionType.MUSIC_AI);

console.log('Music generated!', result);

🔧 API Reference

Client Initialization

const client = new MusicGPTClient(apiKey, logLevel);

Parameters:

• apiKey (string): Your MusicGPT API key
• logLevel (optional): 'DEBUG' | 'INFO' | 'ERROR' (default: 'DEBUG')

Core Methods

🎵 Music Generation

const response = await client.music_ai({
  prompt: 'A relaxing jazz melody',
  music_style: 'jazz',
  lyrics: 'Optional lyrics here',
  make_instrumental: false,
  webhook_url: 'https://your-webhook.com/callback',
});

🎤 Voice Conversion

const response = await client.voice_changer({
  voice_id: 'voice-id-here',
  audio_url: 'https://example.com/audio.mp3',
  pitch: 0,
  remove_background: true,
});

🎧 Audio Cover Generation

const response = await client.cover({
  voice_id: 'voice-id-here',
  audio_url: 'https://example.com/song.mp3',
  pitch: 2,
});

🗣️ Text to Speech

const response = await client.text_to_speech({
  text: 'Hello, this is a test message',
  voice_id: 'optional-voice-id',
  language: 'en',
  gender: 'f',
});

🔍 Voice Search

// Get all voices
const voices = await client.getAllVoices(0, 20);

// Search for specific voices
const femaleVoices = await client.searchVoices('female', 0, 10);

⏳ Wait for Completion

const result = await client.waitForCompletion(
  taskId,
  conversionType,
  3000, // Poll interval (ms)
  300000, // Timeout (ms)
  1 // Verbose level (0-2)
);

Audio Processing

🔊 Denoise Audio

const response = await client.denoise({
  audio_url: 'https://example.com/noisy-audio.mp3',
});

📝 Audio Transcription

const response = await client.audio_transcription({
  audio_url: 'https://example.com/speech.mp3',
  language: 'en',
  translate: false,
  word_timestamps: true,
});

🔄 File Format Conversion

const response = await client.file_conversion({
  audio_url: 'https://example.com/audio.wav',
  target_sampling_rate: 44100,
  target_format: 'mp3',
  target_bit_depth: 16,
});

🎵 Music Analysis

// Extract key and BPM
const keyBpm = await client.extract_key_bmp({
  audio_url: 'https://example.com/song.mp3',
});

// Extract stems (vocals, instruments, etc.)
const stems = await client.extraction({
  audio_url: 'https://example.com/song.mp3',
  stems: ['vocals', 'instrumental', 'drums'],
});

✂️ Audio Cutter

const response = await client.audio_cutter({
  audio_url: 'https://example.com/audio.mp3',
  start_time: 10.5,
  end_time: 30.0,
  output_extension: 'mp3',
});

🎛️ Audio Mastering

const response = await client.audio_mastering({
  audio_url: 'https://example.com/audio.mp3',
  reference_audio_url: 'https://example.com/reference.mp3', // Required
  output_extension: 'wav',
});

⚡ Audio Speed Changer

const response = await client.audio_speed_changer({
  audio_url: 'https://example.com/audio.mp3',
  speed_change_factor: 1.25, // 0.25 to 4.0
  audio_extension: 'mp3',
});

🎹 Audio to MIDI

const response = await client.audio_to_midi({
  audio_url: 'https://example.com/audio.mp3',
  sonify_midi: true,
  save_note_events: false,
});

📈 Rate Limiting

The API has rate limits. The SDK automatically handles:

• Rate limit errors with appropriate exceptions
• Exponential backoff (when implemented)
• Request queuing (planned feature)

⚠️ Error Handling

The SDK provides detailed error information with proper JSON formatting:

try {
  const result = await client.music_ai({
    prompt: 'Generate music',
  });
} catch (error) {
  console.error('API Error:', error.message);
  // Error messages include detailed JSON responses for debugging
}

Common Error Types:

• INTERNAL_SERVER_ERROR - Server-side issues
• 422 Unprocessable Entity - Invalid parameters
• 404 Not Found - Endpoint not available
• Rate limiting errors

📋 Supported Audio Formats

• Input: MP3, WAV, FLAC, OGG, AAC, WEBM
• Output: MP3, WAV, FLAC, OGG, AAC, WEBM
• Sampling Rates: 8kHz to 192kHz
• Bit Depths: 16-bit, 24-bit

Made with ❤️ by the MusicGPT team