Nijivoice

🚀 NijiVoice-MCP

NijiVoice-MCP implements the Model Context Protocol (MCP) version of the にじボイス API. It enables Large Language Models (LLMs) to access the NijiVoice API and easily generate speech from text.

🚀 Quick Start

NijiVoice-MCP is a project under development that allows LLMs to access the NijiVoice API for text-to-speech generation.

⚠️ Important Note

This project is in the development phase and has not been fully verified. Do not use it in a production environment!

✨ Features

• Perform speech synthesis using the NijiVoice API.
• Retrieve information about available voice actors.
• Query the remaining credit balance.
• Connect with LLMs through the MCP interface.

📦 Installation

System Requirements

• Python 3.12 or higher
• NijiVoice API Key
• Internet connection

Installation Steps

# Clone the repository
git clone https://github.com/YOUR_USERNAME/nijivoice-mcp.git
cd nijivoice-mcp

# Install dependencies
pip install -e .

Environment Configuration

1. Create a .env file and add the following content:

NIJIVOICE_API_KEY=Your_API_Key
DEFAULT_VOICE_ACTOR_ID=Default_voice_actor_ID (optional)
LOG_LEVEL=INFO

2. You can obtain the API Key on the NijiVoice official website.

Start the MCP Server

python main.py

After starting, the MCP server will start listening for requests from the LLM.

💻 Usage Examples

Basic Usage

LLM calls the MCP service:

# Get the list of available voice actors
actors = await tools.call("actor/get_voice_actors")

# Generate voice
voice_result = await tools.call("voice/generate_voice", {
    "script": "你好，这是 NijiVoice 的测试。",
    "voice_actor_id": "90031163-c497-44f3-a8a6-e45e4d0cb8f6",  # Voice actor ID
    "speed": 1.0,
    "emotional_level": 0.5,
    "sound_duration": 0.1,
    "format": "mp3"
})

# Query the remaining credit balance
balance = await tools.call("credit/get_credit_balance")

📚 Documentation

Available MCP Tools

Voice Actor Related Tools

• actor/get_voice_actors(): Retrieve information about all available voice actors.

Speech Synthesis Tools

• voice/generate_voice(): Convert text to speech.

Project Structure

nijivoice-mcp/
├── niji_mcp/
│   ├── __init__.py
│   ├── api.py          # API client implementation
│   ├── exceptions.py   # Exception definitions
│   └── models.py       # Data models
├── server.py           # MCP server implementation
└── utils/              # Utility functions
    ├── __init__.py
    ├── audio_utils.py   # Audio processing tools
    ├── error_handling.py# Error handling logic
    └── retry_logic.py   # Retry mechanism

Testing

# Run tests
pytest

Notes and Known Issues

1. Project under development: Features and specifications may change at any time.
2. Risk of NijiVoice API changes: If the NijiVoice API is updated, this project may become incompatible.
3. Insufficient error handling: Error handling in some scenarios is not yet perfect.
4. Timeout for long text generation: Timeout may occur when processing long texts.
5. Incomplete support for response formats: Not all API response formats of NijiVoice are fully supported.

Troubleshooting

If you encounter issues:

1. Check the log file nijivoice_mcp.log.
2. Confirm that the API Key is correctly configured.
3. Check the network connection status.
4. Check the operation status of the NijiVoice API.

📄 License

This project is licensed under the MIT License.

Acknowledgments

Thanks for referring to FastMCP and Model Context Protocol, and using the speech synthesis technology of NijiVoice.