Narakeet MCP Server

An MCP (Model Context Protocol) server for the Narakeet Text-to-Speech API. This server allows LLMs to generate audio files (MP3/M4A/WAV) from text scripts, including bulk/batch processing.

Features

• Singular Text-to-Speech: Convert short scripts directly to audio (Base64 data URI).
• Bulk/Long Text-to-Speech: Convert long scripts (up to 100KB) via asynchronous polling.
• Batch Processing: Convert multiple text segments at once into a ZIP file.
• Voice Discovery: List all available Narakeet voices with filtering by language.

Installation

Prerequisites

• A Narakeet API Key.

Install via NPM

npm install -g narakeet-mcp-server

Configuration

To use this with an MCP client (like Claude Desktop), add the following to your configuration:

{
  "mcpServers": {
    "narakeet": {
      "command": "npx",
      "args": ["-y", "narakeet-mcp-server"],
      "env": {
        "NARAKEET_API_KEY": "your_api_key_here"
      }
    }
  }
}

Tools

text_to_speech

Convert a single piece of text to audio.

• text: (string) The script to convert.
• voice: (string) The voice identifier (e.g., 'brian').
• format: (string, optional) 'mp3', 'm4v', or 'wav'. Default: 'mp3'.
• speed: (number, optional) Speed multiplier (0.1 - 5.0). Default: 1.0.
• volume: (string, optional) 'soft', 'medium', 'loud'. Default: 'medium'.

batch_text_to_speech

Convert multiple segments into a ZIP of audio files.

• texts: (array of strings) Segments to convert.
• voice: (string) Voice identifier.
• format, speed, volume: (optional) Same as above.

list_voices

List available voices.

• language: (string, optional) Filter by language name or code (e.g., 'English', 'en-US').

License

MIT