MCP TurboClient

What is MCP TurboClient?

MCP TurboClient is a FastAPI-based web application that provides a chat interface for connecting to multiple MCP (Model Context Protocol) servers and processing messages through OpenAI’s API. It runs as a Python web server and is intended for developers who need a browser-based way to interact with MCP servers.

How to use MCP TurboClient?

1. Clone the repository and install dependencies with uv pip install -r requirements.txt.
2. Set the MODEL_NAME environment variable (default gpt-3.5-turbo) and create a config.json file listing your MCP servers.
3. Initialize the SQLite database for storing the OpenAI API key, then start the server with uvicorn main:app --reload.
4. Open http://localhost:8000, enter your OpenAI API key via the web interface, and use the chat window to send messages that are routed through the configured MCP servers and OpenAI.

Key features of MCP TurboClient

• Connect to multiple MCP servers simultaneously from one interface.
• Real-time chat via WebSocket for message processing.
• Web-based UI built with HTML, CSS, Bootstrap, and Animate.css.
• Stores OpenAI API license key securely in a SQLite database.
• Supports configuration of MCP servers via a JSON config file.

Use cases of MCP TurboClient

• Developers wanting a quick web UI for testing multiple MCP servers.
• Connecting AI assistants (backed by OpenAI) to external MCP-based tools.
• Prototyping multi‑server workflows without writing a full client application.

FAQ from MCP TurboClient

What does MCP TurboClient do?

It provides a browser‑based chat interface that sends user messages to one or more MCP servers and then processes the results using OpenAI’s API, returning the final response to the user.

How do I connect MCP servers?

Create a config.json file in the root directory with the mcpServers key, specifying each server’s command and arguments. The app reads this file on startup.

Which OpenAI models are supported?

You can set any model name via the MODEL_NAME environment variable (e.g., gpt-4-turbo, gpt-3.5-turbo). The app passes the chosen model to OpenAI’s API.

Is MCP TurboClient free to use?

The software itself is released under the MIT license and is free to use. You must provide your own OpenAI API key, which may incur usage costs.

What are the main limitations?

The README does not mention specific limitations. The app requires a Python environment and a running FastAPI server; all data is stored locally in an SQLite database.