perplexity-mcp logo

perplexity-mcp

by jsonallen

A Model Context Protocol (MCP) server that provides web search functionality using Perplexity AI's API. It is designed to work with the Anthropic Claude desktop client.

View on GitHub

Last updated: N/A

What is perplexity-mcp?

The perplexity-mcp server is an MCP (Model Context Protocol) server that integrates Perplexity AI's web search capabilities into applications like the Anthropic Claude desktop client. It allows users to perform web searches directly from within these applications using natural language prompts.

How to use perplexity-mcp?

To use the server, you need to install it via Smithery or manually using UV (a fast Python package and project manager). You also need to set the PERPLEXITY_API_KEY environment variable with your Perplexity AI API key. After installation, configure your Cursor or Claude Desktop client to use the server by adding it to the MCP configuration file.

Key features of perplexity-mcp

  • Integrates Perplexity AI web search with Claude and Cursor.

  • Supports filtering search results by recency (day, week, month, year).

  • Provides a single prompt: perplexity_search_web.

  • Implements a tool: perplexity_search_web.

  • Configurable Perplexity model selection.

Use cases of perplexity-mcp

  • Searching the web for recent news or information within Claude or Cursor.

  • Gathering context from the web for AI models.

  • Integrating web search capabilities into AI-powered workflows.

  • Researching specific topics using natural language queries.

FAQ from perplexity-mcp

What is an MCP server?

MCP stands for Model Context Protocol. An MCP server facilitates communication and data exchange between different AI models and applications.

Where do I get a Perplexity AI API key?

You can obtain an API key from the Perplexity AI website (perplexity.ai).

What Perplexity models are supported?

The server supports various Perplexity models, including sonar-deep-research, sonar-reasoning-pro, sonar-reasoning, sonar-pro, sonar and r1-1776. The default model is sonar.

How do I filter search results by recency?

Use the 'recency' argument in your search query, specifying 'day', 'week', 'month', or 'year'.

Where can I find an updated list of available models?

An updated list of models is available at https://docs.perplexity.ai/guides/model-cards