Deepseek Thinker MCP Server

What is Deepseek Thinker MCP Server about?

Delivers the thought process and reasoning outputs of the Deepseek model to any MCP‑enabled AI client such as Claude Desktop. It can operate against the official Deepseek API or a locally hosted Ollama instance.

How to use Deepseek Thinker MCP Server?

1. Install the package via npx (no global install required).
2. Configure environment variables for the desired mode:
   • OpenAI API mode: API_KEY (your Deepseek/OpenAI key) and BASE_URL (API endpoint).
   • Ollama mode: set USE_OLLAMA=true.
3. Add the server definition to the client’s claude_desktop_config.json (or equivalent) using the provided command, args, and env sections.
4. Start the server (e.g., npx -y deepseek-thinker-mcp) and invoke the get-deepseek-thinker tool from the client.

Key Features of Deepseek Thinker MCP Server

• Dual mode support – works with the remote Deepseek API (OpenAI‑compatible) and with a local Ollama model.
• Focused reasoning – captures and returns the model’s internal chain‑of‑thought, exposing structured reasoning to the client.
• TypeScript implementation – strong typing and validation via Zod.
• Simple integration – single‑command launch via npx and straightforward JSON configuration for AI clients.

Use Cases of Deepseek Thinker MCP Server

• Enhancing AI assistants with transparent chain‑of‑thought explanations.
• Debugging or auditing model decisions by exposing intermediate reasoning steps.
• Building custom AI workflows that need both answer generation and the underlying rationale.
• Running locally with Ollama for privacy‑sensitive environments while still accessing Deepseek‑style reasoning.

FAQ from Deepseek Thinker MCP Server

• Why do I see “MCP error -32001: Request timed out”? The Deepseek API response took too long or the reasoning output exceeded the server’s timeout limit.
• Do I need both API_KEY and BASE_URL? For OpenAI API mode both are required; BASE_URL points to the Deepseek endpoint. In Ollama mode only USE_OLLAMA=true is needed.
• Can I run the server without installing globally? Yes, launch it with npx -y deepseek-thinker-mcp as shown in the configuration examples.
• What input does the get-deepseek-thinker tool expect? A single string parameter originPrompt containing the user’s original query.

Deepseek Thinker MCP Server

A MCP (Model Context Protocol) provider Deepseek reasoning content to MCP-enabled AI Clients, like Claude Desktop. Supports access to Deepseek's thought processes from the Deepseek API service or from a local Ollama server.

Core Features

• 🤖 Dual Mode Support

  • OpenAI API mode support
  • Ollama local mode support

• 🎯 Focused Reasoning

  • Captures Deepseek's thinking process
  • Provides reasoning output

Available Tools

get-deepseek-thinker

• Description: Perform reasoning using the Deepseek model
• Input Parameters:
  • originPrompt (string): User's original prompt
• Returns: Structured text response containing the reasoning process

Environment Configuration

OpenAI API Mode

Set the following environment variables:

API_KEY=<Your OpenAI API Key>
BASE_URL=<API Base URL>

Ollama Mode

Set the following environment variable:

USE_OLLAMA=true

Usage

Integration with AI Client, like Claude Desktop

Add the following configuration to your claude_desktop_config.json:

{
  "mcpServers": {
    "deepseek-thinker": {
      "command": "npx",
      "args": [
        "-y",
        "deepseek-thinker-mcp"
      ],
      "env": {
        "API_KEY": "<Your API Key>",
        "BASE_URL": "<Your Base URL>"
      }
    }
  }
}

Using Ollama Mode

{
  "mcpServers": {
    "deepseek-thinker": {
      "command": "npx",
      "args": [
        "-y",
        "deepseek-thinker-mcp"
      ],
      "env": {
        "USE_OLLAMA": "true"
      }
    }
  }
}

Local Server Configuration

{
  "mcpServers": {
    "deepseek-thinker": {
      "command": "node",
      "args": [
        "/your-path/deepseek-thinker-mcp/build/index.js"
      ],
      "env": {
        "API_KEY": "<Your API Key>",
        "BASE_URL": "<Your Base URL>"
      }
    }
  }
}

Development Setup

# Install dependencies
npm install

# Build project
npm run build

# Run service
node build/index.js

FAQ

Response like this: "MCP error -32001: Request timed out"

This error occurs when the Deepseek API response is too slow or when the reasoning content output is too long, causing the MCP server to timeout.

Tech Stack

• TypeScript
• @modelcontextprotocol/sdk
• OpenAI API
• Ollama
• Zod (parameter validation)

License

This project is licensed under the MIT License. See the LICENSE file for details.