MCP Radar

Perplexity MCP Server

The Perplexity MCP Server lets your AI call Perplexity's web-connected Q&A — you get real-time answers with source citations, not the model guessing on its own.

For example, "check what breaking changes the latest version of this library has" — the AI uses Perplexity to look it up online and return with citations. This is Perplexity's official server.

See this server's health data →

What it does for you

Plugs Perplexity's web-grounded Q&A into your AI — real-time search with cited answers, for research and fact-checking.

Capabilities

  • Real-time web Q&A
  • Return cited results

Once installed, try asking your AI

Say this directly in Claude / Cursor or any client with this server connected

  • Look up the latest best practices for this tech, with sources

Install / Connect

# No npm package — install from source.
# See the repo's README for the exact command:
# https://github.com/perplexityai/modelcontextprotocol

Follow the official README; env and args vary per server.

Why use it

Good for: developers adding real-time, traceable information to the AI.

Official implementation with a high star count; full health data is shown below.

Similar / alternative servers

Frequently Asked Questions

Are the answers trustworthy?

Perplexity returns answers with source citations you can click through to verify.

Does it need an API key?

Yes, a Perplexity API key, billed by usage.