AUTOMATIONSWITCH
CommunitySearch

DuckDuckGo Search MCP

by nickclyde

Community-built DuckDuckGo Search MCP server with web search, content fetching, and intelligent text extraction. Built-in rate limiting protects against DuckDuckGo blocking. SafeSearch and region configuration via env vars.

1,043·2 tools·Released APR 2025·MIT
uvx duckduckgo-mcp-server
Share:

DuckDuckGo Search MCP is the community-built privacy-first search server, and the architectural priorities show. Rate limiting is built in for both search and content fetching, keeping the agent within DuckDuckGo's tolerance and the operator's IP outside the block list. Search results are formatted specifically for LLM consumption, in clean text form. Content fetching includes intelligent text extraction so when the agent follows a search result to read a page, the response arrives as parsed markdown. The standout feature is configurability with anonymous access: SafeSearch level (`STRICT`, `MODERATE`, `OFF`) and default region (`us-en`, `cn-zh`, `jp-ja`, etc.) configure via env vars, no API key required. The trade-off is DuckDuckGo's results quality: for general web search, DuckDuckGo's index is solid but smaller than Google's, so research-intensive workflows may want a paid search MCP alongside this one. For agents that need fast, privacy-respecting search out of the box, this is the editorial pick.

Reviewed by M. Nouriel · APR 2026

INSTALL THIS SERVER

{ "mcpServers": { "ddg-search": { "command": "uvx", "args": [ "duckduckgo-mcp-server" ], "env": { "DDG_SAFE_SEARCH": "MODERATE" } } } }
PrereqRequires uv. No API key required. Optional env vars: DDG_SAFE_SEARCH (STRICT/MODERATE/OFF), DDG_REGION (us-en, jp-ja, etc.). Path: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS).
{ "mcpServers": { "ddg-search": { "command": "uvx", "args": [ "duckduckgo-mcp-server" ], "env": { "DDG_SAFE_SEARCH": "MODERATE" } } } }
{ "mcpServers": { "ddg-search": { "command": "uvx", "args": [ "duckduckgo-mcp-server" ], "env": { "DDG_SAFE_SEARCH": "MODERATE" } } } }
{ "mcpServers": { "ddg-search": { "command": "uvx", "args": [ "duckduckgo-mcp-server" ], "env": { "DDG_SAFE_SEARCH": "MODERATE" } } } }
{ "mcpServers": { "ddg-search": { "command": "uvx", "args": [ "duckduckgo-mcp-server" ], "env": { "DDG_SAFE_SEARCH": "MODERATE" } } } }

2 TOOLS AVAILABLE

search
Run a DuckDuckGo web search with rate limiting and LLM-friendly result formatting
Read
fetch_content
Retrieve a webpage and extract clean markdown content
Read

OUR ASSESSMENT

Strengths
  • No API key required; runs out of the box via `uvx duckduckgo-mcp-server`.
  • Built-in rate limiting for both search and content fetching protects against DuckDuckGo blocking the operator's IP.
  • Content fetching includes intelligent text extraction; the agent receives parsed markdown in place of raw HTML.
  • Configurable SafeSearch (`STRICT`, `MODERATE`, `OFF`) and default region via env vars (`DDG_SAFE_SEARCH`, `DDG_REGION`).
Weaknesses
  • DuckDuckGo's search index is smaller than Google's; high-volume research workflows benefit from a paid search MCP alongside this one.
  • Community-maintained; bus factor sits with a single maintainer.
  • Rate limits are conservative by design; teams running high-frequency search loops will hit them.
Security Notes

The server runs locally and queries DuckDuckGo's public search endpoint with no authentication required. SafeSearch defaults to `MODERATE`; teams handling sensitive content can tighten to `STRICT` via the `DDG_SAFE_SEARCH` env var. Content fetching follows search-result links; for queries that may return pages with sensitive or untrusted content, consider running the server in a sandboxed environment. Rate limiting is internal to the server and operates per-process, so multiple parallel server instances can collectively exceed DuckDuckGo's tolerance.

Best For

Teams who want privacy-respecting web search inside agent workflows with anonymous access. Pair with a paid search MCP for high-volume research.

TECHNICAL DETAILS

Language
python
Transport
stdio
Clients
Claude DesktopClaude CodeCursorVS CodeWindsurf
License
MIT
npm
duckduckgo-mcp-server
Last Release
v0.3.0APR 26, 2026
First Released
APR 22, 2025

ADOPTION METRICS

// GitHub Stars
1,043

// Reading this1k stars in twelve months reflect adoption from teams who want privacy-first search inside agent workflows. Star traction places this as the leading open-source search MCP.

// Popularity Rank
#16
Globally · #1 in Search

// Reading thisTop-ranked search MCP and the only zero-credentials search server in the directory. The built-in rate limiting and LLM-friendly output set it apart from community alternatives.

SOURCES & VERIFICATION

We don't take any single directory's word for it. Before scoring, we cross-reference 5 public MCP sources, install the server ourselves against the clients we cover, and record when we last re-verified.

01
Discovered
Manual submission
First indexed APR 28, 2026
02
Cross-referenced
5 directories
PulseMCP, Official MCP Registry, Awesome MCP Servers, Smithery, MCP.so
03
Verified against
Claude Desktop, Cursor, VS Code, Windsurf
Installed and tested across clients
04
Last re-checked
APR 28, 2026
Weekly re-verification
// How other directories see it

The same server, 5 different lenses. We reconcile these signals into our editorial score, which is why our number sometimes diverges from a directory-aggregate star count.

SourceTheir ratingTheir star countTheir downloadsLast synced
AutomationSwitch This page4editorial1,043APR 28, 2026
PulseMCP— unratedunavailableunavailableAPR 28, 2026
Official MCP Registry— unratedunavailableunavailableAPR 28, 2026
Awesome MCP Servers— unratedunavailableunavailableAPR 28, 2026
Smithery— unratedunavailableunavailableAPR 28, 2026
MCP.so— unratedunavailableunavailableAPR 28, 2026

// Counts are directory-reported; we don't adjust them. Discrepancies usually come from different snapshot times or star-caching.

// Get in touch

DISCUSS YOUR
MCP REQUIREMENTS.

Evaluating a server, scoping an internal deployment, or working out whether MCP is the right fit at all. Start the conversation and we will point you at the right piece of the ecosystem.

Discuss Your MCP Requirements →