Deepseek R1

Provides an MCP server to run the Deepseek R1 model with a 8192-token context window for Claude Desktop.

typescript

1

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "mcp-mirror-66julienmartin_mcp-server-deepseek_r1": {
      "command": "node",
      "args": [
        "/path/to/deepseek-r1-mcp/build/index.js"
      ],
      "env": {
        "DEEPSEEK_API_KEY": "your-api-key"
      }
    }
  }
}

You run an MCP server that hosts the Deepseek R1 language model, enabling Claude Desktop and other MCP clients to generate context-aware responses with a large 8192-token window. This server abstracts model interaction, argument handling, and error reporting, so you can focus on building prompts and workflows rather than low-level RPC details.

How to use

Start the MCP server as a local process and configure your MCP client to connect via the provided stdio channel. You’ll send prompts and receive model completions through the MCP protocol, with the server handling token limits, temperature, and max token settings. Use it to integrate Deepseek R1 into your Claude Desktop setup or other MCP-enabled clients, enabling reasoning tasks, data-driven prompts, and lengthy contextual conversations.

How to install

Prerequisites you need on your machine before proceeding:

Node.js (v18 or higher) and npm
Claude Desktop (for MCP client usage)
A Deepseek API key to authenticate requests to the model

Step-by-step commands

Clone the repository and install dependencies, then build the project and prepare for local execution.

# Clone the MCP server repository
git clone https://github.com/66julienmartin/MCP-server-Deepseek_R1.git
cd deepseek-r1-mcp
npm install

# Set up environment
cp .env.example .env  # Then add your API key

# Build for production
npm run build

Configure the MCP client (Claude Desktop)

Create the MCP connection configuration so Claude Desktop can communicate with the local server. Use the standard stdio approach shown here.

{
  "mcpServers": {
    "deepseek_r1": {
      "command": "node",
      "args": ["/path/to/deepseek-r1-mcp/build/index.js"],
      "env": {
        "DEEPSEEK_API_KEY": "your-api-key"
      }
    }
  }
}

Run the server locally

Start the server with the runtime command and verify it’s listening for MCP connections. Ensure the API key environment variable is correctly set in the environment you launch from.

Additional configuration

Environment and model selection can be adjusted to your needs. The default model is the Deepseek R1. If you need to switch to a different Deepseek model, follow the model selector notes in the source and adjust the build accordingly.

Configuration

The server configuration requires your API key to authenticate requests to the Deepseek service. You provide this key through an environment variable named DEEPSEEK_API_KEY.

Example environment setup and startup flow are shown in the installation steps above. The configuration you embed in Claude Desktop maps a named MCP server to a local Node.js process that runs the built index.

Development

For development, you can run in watch mode to automatically rebuild on changes and then restart the local MCP server as needed.