PluginBench
MCP Server
Maintained
MIT

DeepSeek MCP Server MCP Server

io.github.DMontgomery40/deepseek

Official MCP server exposing DeepSeek V4 chat, FIM completion, model listing, and balance endpoints.

What is the DeepSeek MCP Server MCP server?

DeepSeek MCP Server is a Model Context Protocol server that connects AI agents like Claude and Cursor to DeepSeek's V4 API, exposing chat completion, FIM completion, model listing, and account balance endpoints. It supports both local stdio (npx/Docker) installation with your own DeepSeek API key, and a hosted remote HTTP endpoint with bearer-token auth.

DeepSeek MCP Server gives AI coding assistants and agents (Claude Code, Cursor, Codex CLI, etc.) direct access to DeepSeek's V4 API through the Model Context Protocol. It supports chat completions (with thinking/reasoning modes, JSON output, function tools, streaming, and conversation memory), fill-in-the-middle code completion via deepseek-v4-pro, live model listing, and account balance checks. It can run locally via npx or Docker using your own DeepSeek API key, or connect to a hosted remote endpoint with bearer-token authentication.

How to install DeepSeek MCP Server

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
Environment / auth
  • DEEPSEEK_API_KEY
    required
    secret

    DeepSeek API key used for upstream DeepSeek requests

~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "deepseek": {
      "command": "npx",
      "args": [
        "-y",
        "deepseek-mcp-server"
      ],
      "env": {
        "DEEPSEEK_API_KEY": "<YOUR_DEEPSEEK_API_KEY>"
      }
    }
  }
}

Tools & capabilities

Tools this server exposes to the agent.

  • chat_completion — DeepSeek V4 chat, defaults to deepseek-v4-flash. Supports thinking mode (enabled/disabled), reasoning_effort (high/max), JSON output, function tools, logprobs, streaming, and conversation memory.
  • completion — DeepSeek V4 Pro FIM (fill-in-the-middle) completion, defaults to deepseek-v4-pro.
  • list_models — Reads the live DeepSeek model list.
  • get_user_balance — Reads DeepSeek account balance and availability.
  • reset_conversation — Clears an in-memory conversation.
  • list_conversations — Lists in-memory conversation IDs.

Use cases

  • Add DeepSeek V4 chat capabilities (including reasoning/thinking mode) to an AI coding agent
  • Perform code fill-in-the-middle completions using deepseek-v4-pro
  • Check available DeepSeek models programmatically
  • Monitor DeepSeek account balance from within an agent workflow
  • Maintain and reset multi-turn conversation state during agent sessions

DeepSeek MCP Server MCP server FAQ

What does the DeepSeek MCP server do?

It exposes DeepSeek's V4 API (chat completions, FIM completions, model listing, and user balance) as MCP tools so agents like Claude Code and Cursor can call DeepSeek directly.

Is the DeepSeek MCP server free to use?

The server itself is open source (MIT license) and free, but you need a DeepSeek API key, and usage is billed by DeepSeek according to their API pricing/balance.

How do I install it in Cursor?

You can add it to Cursor's mcp.json pointing at the hosted remote endpoint with an Authorization bearer token header, or run it locally via `npx -y deepseek-mcp-server` with DEEPSEEK_API_KEY set and configure Cursor to use the stdio transport.

How do I install it in Claude Code?

Run `claude mcp add --transport http deepseek https://deepseek-mcp.ragweld.com/mcp --header "Authorization: Bearer $DEEPSEEK_MCP_AUTH_TOKEN"` after setting the DEEPSEEK_MCP_AUTH_TOKEN environment variable, or configure it locally via stdio with a DeepSeek API key.

What authentication does it require?

For local/stdio use it requires a DEEPSEEK_API_KEY environment variable. For the hosted remote endpoint, it requires a bearer token passed via the Authorization header (DEEPSEEK_MCP_AUTH_TOKEN).

Can I self-host this server?

Yes, it can be run locally via npx (`npx -y deepseek-mcp-server`) or via the published Docker image, using your own DeepSeek API key and optional environment configuration for base URL, timeouts, default model, and transport settings.

README (reference)

Source of truth, from the repository.

DeepSeek MCP Server

<p align="center"> <a href="https://github.com/deepseek-ai/awesome-deepseek-integration"><img alt="DeepSeek Official List" src="https://img.shields.io/badge/DeepSeek%20Official%20List-Linked-0A66FF?logo=github&logoColor=white" /></a> <a href="https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.DMontgomery40/deepseek"><img alt="Official MCP Registry" src="https://img.shields.io/badge/MCP%20Registry-Official%20Active-0A66FF" /></a> <a href="https://www.npmjs.com/package/deepseek-mcp-server"><img alt="npm version" src="https://img.shields.io/npm/v/deepseek-mcp-server?logo=npm" /></a> <a href="https://www.npmjs.com/package/deepseek-mcp-server"><img alt="npm downloads" src="https://img.shields.io/npm/dm/deepseek-mcp-server?logo=npm" /></a> <a href="https://github.com/DMontgomery40/deepseek-mcp-server/blob/main/server.json"><img alt="OCI package" src="https://img.shields.io/badge/OCI-docker.io%2Fdmontgomery40%2Fdeepseek--mcp--server%3A0.5.0-2496ED?logo=docker&logoColor=white" /></a> <a href="https://github.com/DMontgomery40/deepseek-mcp-server"><img alt="GitHub stars" src="https://img.shields.io/github/stars/DMontgomery40/deepseek-mcp-server?logo=github" /></a> <a href="https://glama.ai/mcp/servers/asht4rqltn"><img alt="Glama MCP Listing" src="https://img.shields.io/badge/Glama-MCP%20Listing-7B61FF" /></a> <a href="https://spark.entire.vc/assets/vb-deepseek-mcp-server?utm_source=github&utm_medium=readme"><img alt="Listed on Spark" src="https://spark.entire.vc/badges/listed.svg" /></a> <a href="https://spark.entire.vc/assets/vb-deepseek-mcp-server?utm_source=github&utm_medium=readme"><img alt="Install via Spark" src="https://spark.entire.vc/badges/vb-deepseek-mcp-server/install.svg" /></a> </p>

Model Context Protocol server for the current DeepSeek V4 API.

As of April 24, 2026, DeepSeek's public API reference documents:

  • POST /chat/completions with deepseek-v4-flash and deepseek-v4-pro
  • POST /beta/completions for V4 Pro FIM completion
  • GET /models
  • GET /user/balance

This server exposes only those documented API surfaces. It does not ship a V4 monitor, speculative image/video/upload tools, or automatic model substitution.

Tools

  • chat_completion: DeepSeek V4 chat. Defaults to deepseek-v4-flash. Supports thinking: { "type": "enabled" | "disabled" }, reasoning_effort: "high" | "max", JSON output, function tools, logprobs, streaming, and conversation memory.
  • completion: DeepSeek V4 Pro FIM completion. Defaults to deepseek-v4-pro.
  • list_models: Reads the live DeepSeek model list.
  • get_user_balance: Reads account balance and availability.
  • reset_conversation: Clears an in-memory conversation.
  • list_conversations: Lists in-memory conversation IDs.

Hosted Remote

  • URL: https://deepseek-mcp.ragweld.com/mcp
  • Auth: Authorization: Bearer <token>

Codex CLI:

export DEEPSEEK_MCP_AUTH_TOKEN="REPLACE_WITH_TOKEN"
codex mcp add deepseek --url https://deepseek-mcp.ragweld.com/mcp --bearer-token-env-var DEEPSEEK_MCP_AUTH_TOKEN

Claude Code:

export DEEPSEEK_MCP_AUTH_TOKEN="REPLACE_WITH_TOKEN"
claude mcp add --transport http deepseek https://deepseek-mcp.ragweld.com/mcp --header "Authorization: Bearer $DEEPSEEK_MCP_AUTH_TOKEN"

Cursor:

node -e 'const fs=require("fs"),p=process.env.HOME+"/.cursor/mcp.json";let j={mcpServers:{}};try{j=JSON.parse(fs.readFileSync(p,"utf8"))}catch{};j.mcpServers={...(j.mcpServers||{}),deepseek:{url:"https://deepseek-mcp.ragweld.com/mcp",headers:{Authorization:"Bearer ${env:DEEPSEEK_MCP_AUTH_TOKEN}"}}};fs.mkdirSync(process.env.HOME+"/.cursor",{recursive:true});fs.writeFileSync(p,JSON.stringify(j,null,2));'

Local Stdio

DEEPSEEK_API_KEY="REPLACE_WITH_DEEPSEEK_KEY" npx -y deepseek-mcp-server

Docker:

docker pull docker.io/dmontgomery40/deepseek-mcp-server:0.5.0
docker run --rm -i -e DEEPSEEK_API_KEY="REPLACE_WITH_DEEPSEEK_KEY" docker.io/dmontgomery40/deepseek-mcp-server:0.5.0

Environment

Required:

DEEPSEEK_API_KEY=your-api-key

Optional:

DEEPSEEK_BASE_URL=https://api.deepseek.com
DEEPSEEK_REQUEST_TIMEOUT_MS=120000
DEEPSEEK_DEFAULT_MODEL=deepseek-v4-flash
MCP_TRANSPORT=stdio
MCP_HTTP_HOST=127.0.0.1
MCP_HTTP_PORT=3001
MCP_HTTP_PATH=/mcp
MCP_HTTP_STATEFUL_SESSION=false
CONVERSATION_MAX_MESSAGES=200

Verification

npm run build
npm test
DEEPSEEK_API_KEY="REPLACE_WITH_DEEPSEEK_KEY" npm run test:live
DEEPSEEK_MCP_AUTH_TOKEN="REPLACE_WITH_TOKEN" npm run test:remote

The live smoke test performs real DeepSeek requests for model listing, balance, non-thinking chat, thinking streaming chat with reasoning_content, FIM completion, and MCP tool calls.

Registry Identity

  • MCP Registry name: io.github.DMontgomery40/deepseek
  • npm package: deepseek-mcp-server
  • OCI package: docker.io/dmontgomery40/deepseek-mcp-server:0.5.0

Official References

License

MIT

Related MCP servers

Elixir/OTP runtime rendering one TEA app to terminal, browser, SSH, and MCP agent tools.

76
Elixir
MIT
View repository →

Search official UK tenders and create source-linked, deterministic supplier-fit assessments.

SCScalpStream logo

Pay-per-call data over x402: air quality, border waits, fuel prices, market research.

0
Go
MIT
View repository →

Independent static verification for exact immutable public GitHub commits.

0
JavaScript
MIT
View repository →

AI sports predictions for NBA, NFL, NHL & MLB — picks, cards, track record, schedule

View repository →

Query Airbrake errors, notices, deploys, and project stats from MCP clients