Kokoro Tts Mcp

@mberg

About Kokoro Tts Mcp

Kokoro Text to Speech (TTS) MCP Server

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "kokoro-tts-mcp": {
      "command": "python",
      "args": [
        "mcp_client.py",
        "--text",
        "Hello, world!"
      ]
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Kokoro Tts Mcp?

Kokoro Tts Mcp is a Model Context Protocol server that converts text to speech using the Kokoro TTS model. It generates .mp3 files that can be stored locally and optionally uploaded to Amazon S3. It is designed for developers who need programmatic TTS generation as part of an AI agent or automation pipeline.

How to use Kokoro Tts Mcp?

Clone the repository, download the Kokoro ONNX weights (kokoro-v1.0.onnx and voices-v1.0.bin), and install ffmpeg. Add the MCP configuration with required environment variables, then run uv run mcp-tts.py. Use the included client script mcp_client.py with --text or --file to send TTS requests.

Key features of Kokoro Tts Mcp

Generates .mp3 files from text input
Optional S3 upload for cloud storage
Configurable voice, speed, and language
Local MP3 retention and automatic cleanup
Supports text input from file or argument
Works with the Kokoro TTS model (ONNX)

Use cases of Kokoro Tts Mcp

Generate audio narration from text content
Automate voiceover creation for applications
Store TTS output in S3 for scalable access
Integrate TTS into AI agent conversations

FAQ from Kokoro Tts Mcp

What dependencies are required to run the server?

Python with uv, ffmpeg (for .wav to .mp3 conversion), and the Kokoro ONNX model weights. The recommended install method is uv run.

How do I configure Amazon S3 uploads?

Set the environment variables AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_S3_BUCKET_NAME, AWS_S3_REGION, and AWS_S3_FOLDER. Enable S3 with S3_ENABLED=true.

How can I change the voice used for TTS?

Set the TTS_VOICE environment variable (default af_heart) or use the --voice flag when calling mcp_client.py.

Where are generated MP3 files stored?

Locally in the folder specified by MP3_FOLDER (default is a mp3 folder in the script directory). They can also be uploaded to S3 if S3_ENABLED is true.

Can I disable S3 upload without changing environment variables?

Yes. Pass the --no-s3 flag to mcp_client.py to disable S3 upload for that request. You can also set S3_ENABLED to false in the environment.

Frequently asked questions

What dependencies are required to run the server?

Python with `uv`, ffmpeg (for .wav to .mp3 conversion), and the Kokoro ONNX model weights. The recommended install method is `uv run`.

How do I configure Amazon S3 uploads?

Set the environment variables `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, `AWS_S3_BUCKET_NAME`, `AWS_S3_REGION`, and `AWS_S3_FOLDER`. Enable S3 with `S3_ENABLED=true`.

How can I change the voice used for TTS?

Set the `TTS_VOICE` environment variable (default `af_heart`) or use the `--voice` flag when calling `mcp_client.py`.

Where are generated MP3 files stored?

Locally in the folder specified by `MP3_FOLDER` (default is a `mp3` folder in the script directory). They can also be uploaded to S3 if `S3_ENABLED` is true.

Can I disable S3 upload without changing environment variables?

Yes. Pass the `--no-s3` flag to `mcp_client.py` to disable S3 upload for that request. You can also set `S3_ENABLED` to `false` in the environment.

Comments

Basic information

More Other MCP servers

MaxKB

1Panel-dev

🔥 MaxKB is an open-source platform for building enterprise-grade agents. 强大易用的开源企业级智能体平台。

22.0K

🚀 Model Context Protocol (MCP) Curriculum for Beginners

microsoft

This open-source curriculum introduces the fundamentals of Model Context Protocol (MCP) through real-world, cross-language examples in .NET, Java, TypeScript, JavaScript, Rust and Python. Designed for developers, it focuses on practical techniques for building modular, scalable,

16.7K