mcpField
H
Hume MCP Server

Hume MCP Server

Local
github.com/HumeAIMedia & Design9Updated 42w ago
Repository

About Hume MCP Server

Hume MCP Server brings Hume's Octave text-to-speech capabilities into AI chats for people working on speech projects. It also supports replaying generated audio and managing voices.

Key features

  1. 01Synthesize speech from text
  2. 02Replay previously generated audio
  3. 03List available voices
  4. 04Save and delete voices

Use cases

  1. 01Generate speech for text
  2. 02Listen to generated audio again
  3. 03Browse available voices
  4. 04Manage a voice library

MCP support

The MCP server exposes speech synthesis, audio replay, and voice-management tools to MCP clients such as Claude Desktop, Cursor, and Windsurf.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "hume": {
      "command": "npx",
      "args": [
        "-y",
        "@humeai/mcp-server"
      ],
      "env": {
        "HUME_API_KEY": "<HUME_API_KEY>"
      }
    }
  }
}
Run commandnpx -y @humeai/mcp-server
Environment variables to set
HUME_API_KEYSecret

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@humeai/mcp-server

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README
README image

Hume MCP Server

Collaborate with AI assistants on your Text-to-Speech projects

📘 Documentation • 💬 Join us on Discord • 🔐 Getting your API Keys

The Hume MCP Server gives you the the ability to use Octave Text to Speech from within an AI Chat, using an MCP Client Application such as Claude Desktop, Cursor, or Windsurf.

Octave TTS is the first text-to-speech system built on LLM intelligence. Octave is a speech-language model that understands what words mean in context, unlocking a new level of expressiveness and nuance. It performs the source text, it doesn't just pronounce it.

See this video for a demonstration of using the MCP Server to narrate a scene from an audiobook.

Quickstart

Click here to add to Cursor:

Install MCP Server

Copy the following into your client's MCP configuration (for example, inside the .mcpServers property of claude_desktop_config.json for Claude Desktop, or of the mcp.json for Cursor).

{
    ...
    "hume": {
        "command": "npx",
        "args": [
            "@humeai/mcp-server"
        ],
        "env": {
            "HUME_API_KEY": "<your_hume_api_key>",
        }
    }
}

Prerequisites

  • An account and API Key from Hume AI
  • Node.js
  • (optional) A command-line audio player
    • ffplay from FFMpeg is recommended, but the server will attempt to detect and use any of several common players.
Available Tools

The server exposes the following MCP tools:

  • tts: Synthesize (and play) speech from text
  • play_previous_audio: Replay previously generated audio
  • list_voices: List available voices
  • save_voice: Save a generated voice to your library
  • delete_voice: Remove a voice from your library
Command Line Options
Options:
  --workdir, -w <path>       Set working directory for audio files (default: system temp)
  --(no-)embedded-audio-mode Enable/disable embedded audio mode (default: false)
  --help, -h                 Show this help message

Evaluation Framework

The project includes a comprehensive evaluation framework that measures how effectively AI agents can utilize the Hume TTS tools across various real-world scenarios.

Environment Variables

  • HUME_API_KEY: Your Hume AI API key (required)
  • WORKDIR: Working directory for audio files (default: system temp directory + "/hume-tts")
  • EMBEDDED_AUDIO_MODE: Enable/disable embedded audio mode (default: false, set to 'true' to enable)
  • ANTHROPIC_API_KEY: Required for running evaluations

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.