FastMCP FastMCP
Sign In
K

Kokoro Speech

Provides text-to-speech capabilities using the Kokoro TTS model, enabling natural-sounding voice out...

5 views
0 installs
Updated Dec 20, 2025
Not audited
AI and Machine Learning Developer Tools
Provides text-to-speech capabilities using the Kokoro TTS model, enabling natural-sounding voice output with customizable playback speed and voice selection through robust error handling and temporary file management.

  1. Open the FastMCP connection interface

    • In the FastMCP UI where you see the server entry and the "Install Now" button, open the connection/config panel for the "speech" MCP server.

  2. Click "Install Now" to open the install / environment dialog

    • Press the ready-made "Install Now" button. When the install dialog or environment-variables panel appears, you will enter the ENV values there.

  3. Add the MCP_DEFAULT_SPEECH_SPEED environment variable

    • Key: MCP_DEFAULT_SPEECH_SPEED
    • Value: a number between 0.5 and 2.0 (default shown in the config: 1.3). Example values: 1.1, 1.3, 1.5.
    • Notes: This is the speed multiplier for TTS. Enter a decimal number (e.g., 1.3).

  4. Add the MCP_DEFAULT_VOICE environment variable

    • Key: MCP_DEFAULT_VOICE
    • Value: voice ID string (default shown: af_bella). Example: af_bella
    • Notes: If you want a different voice, leave this blank temporarily, install, then call the server's list_voices tool to see available voice IDs and update this value.

  5. Save / confirm the environment variables and complete the install

    • Click "Save", "Confirm", or the equivalent action in the FastMCP dialog to apply the ENV values and finish installation.

  6. (Optional) Verify available voices and pick a voice ID

    • If you need a different voice ID, use the server's list_voices tool after installation to list valid voice IDs and update MCP_DEFAULT_VOICE with a chosen ID:
      • Example request (MCP Inspector or raw JSON):
        {"type":"request","id":"1","method":"list_voices","params":}

  7. (Optional) Check model initialization status

    • Use the get_model_status tool to confirm the TTS model is initializing or ready:
      • Example request:
        {"type":"request","id":"1","method":"call_tool","params":{"name":"get_model_status","arguments":}}

  8. No API key required

    • The Speech MCP Server does not require any external API key — only set the two environment variables above as needed.

Quick Start

View on GitHub

More for AI and Machine Learning

View All →

More for Developer Tools

View All →

Similar MCP Servers

O

OpenAI TTS

Enables high-quality voice generation from text using OpenAI's TTS API with customizable voices, formats, and speech parameters for seamless audio playback during conversations.

Entertainment and Media AI and Machine Learning
146
1
V

VOICEPEAK

Integrates with VOICEPEAK engine to generate natural-sounding Japanese speech with customizable narrators, emotions, speed, and pitch parameters for voice content creation and language learning applications.

Entertainment and Media
149
0
1-Click Ready
F

Fish Audio

Integrates with Fish Audio's API to generate high-quality speech from text with configurable voice models, audio formats, and real-time streaming for creating conversational applications and automated content narration.

AI and Machine Learning
282
0

Report Issue

Thank you! Your issue report has been submitted successfully.