Advertise
MCP.so
Sign In

Deepseek Thinker MCP Server

@ruixingshi

Repository

Basic information

Category

AI & Agents

Transports

stdio

Author

ruixingshi

About Deepseek Thinker MCP Server

A MCP provider Deepseek reasoning content to MCP-enabled AI Clients, like Claude Desktop. Supports access to Deepseek's CoT from the Deepseek API service or a local Ollama server.

Deepseek Thinker MCP Server config

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

{
  "mcpServers": {
    "deepseek-thinker-mcp": {
      "command": "node",
      "args": [
        "build/index.js"
      ]
    }
  }
}

Deepseek Thinker MCP Server tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

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

Overview

What is Deepseek Thinker MCP Server?

Deepseek Thinker MCP Server is an MCP (Model Context Protocol) provider that delivers Deepseek reasoning content (including thought processes) to MCP-enabled AI clients, such as Claude Desktop. It supports access either through the Deepseek API service or via a local Ollama server.

How to use Deepseek Thinker MCP Server?

Install and run the server via npx -y deepseek-thinker-mcp or build locally with npm install && npm run build then run node build/index.js. Configure it in your client’s MCP settings, e.g., claude_desktop_config.json, setting environment variables for either OpenAI API mode (API_KEY, BASE_URL) or Ollama mode (USE_OLLAMA=true). Invoke the get-deepseek-thinker tool, providing an originPrompt string, to receive structured reasoning output.

Key features of Deepseek Thinker MCP Server

  • Dual mode support: OpenAI API and Ollama local
  • Captures Deepseek’s thinking process for reasoning output
  • Provides a single get-deepseek-thinker tool for reasoning queries
  • Built with TypeScript, @modelcontextprotocol/sdk, and Zod validation
  • Licensed under MIT

Use cases of Deepseek Thinker MCP Server

  • Integrating Deepseek reasoning directly into Claude Desktop conversations
  • Accessing Deepseek’s thought processes without sending data to the cloud (Ollama mode)
  • Adding reasoning capabilities to any MCP‑enabled AI assistant
  • Experimenting with Deepseek’s reasoning chain for analysis or debugging

Frequently asked questions

1

I see “MCP error -32001: Request timed out”. What does that mean?

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

2

What modes does Deepseek Thinker MCP Server support?

It supports two modes: OpenAI API mode (requires API_KEY and BASE_URL) and Ollama local mode (set USE_OLLAMA=true).

3

How do I switch between modes?

Set the appropriate environment variables in your MCP server configuration. Use `API_KEY` and `BASE_URL` for OpenAI API mode, or set `USE_OLLAMA=true` for Ollama mode.

4

What are the runtime requirements?

Node.js and npm are needed to install dependencies and build the project. The server is a Node.js application built with TypeScript.

5

Where can I find the source and license?

The project is open source under the MIT License; the repository includes the full license file.

Comments

More AI & Agents MCP servers

Ranla

Kurtis Tryber

AI growth agent for software businesses. Ranla ships pages that rank on Google, gets your product named in AI answers, replies on Reddit and sends lifecycle email. Ask how growth is going, have Ranla draft the work and a

FLORA

FLORA is a remote MCP server available at https://agents.flora.ai. The FLORA MCP lets users generate, organize, and iterate on visual assets — hero images, moodboards, lifestyle shots, promo graphics — directly from a co

Poly-Glot AI Workspace

hmoses

1,000+ fully localized prompt templates across 40+ categories. Compare Mode — send the same prompt to up to 9 AI providers (ChatGPT, Claude, Gemini, Grok, Groq, Copilot, Mistral, Perplexity, DuckDuckGo AI) and compare an

F

fhirHydrant

faulkj

Open-source Node.js FHIR MCP server with SMART Backend Services, metadata-aware search/CRUD tools, compact responses, FHIRPath filtering, safe pagination, audit events, and terminology lookup.

14

Reach MCP

Kanbox

Operate a real LinkedIn account from your AI agent. 52 tools for the inbox, invitations, network, profiles, Sales Navigator search in plain language, posts and engagement; six playbooks served as MCP prompts that preview

Advertise