The official MCP server implementation for the Perplexity API Platform, providing AI assistants with real-time web search, reasoning, and research capabilities through Sonar models and the Search API.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/perplexityai/modelcontextprotocol/llms.txt
Use this file to discover all available pages before exploring further.
What is MCP?
The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely access external data sources and tools. The Perplexity MCP Server implements this protocol to bring Perplexity’s powerful search and reasoning capabilities to any MCP-compatible client.Key Features
Real-Time Web Search
Direct access to current web information with ranked search results, snippets, and metadata through the Perplexity Search API.
AI-Powered Research
Deep, comprehensive research using the sonar-deep-research model for thorough analysis and detailed reports with citations.
Advanced Reasoning
Step-by-step logical analysis and problem-solving using the sonar-reasoning-pro model for complex analytical tasks.
Conversational AI
Quick, web-grounded answers using the sonar-pro model for everyday questions and searches with automatic citations.
Available Tools
The Perplexity MCP Server provides four specialized tools, each optimized for different use cases:perplexity_search
Direct web search using the Perplexity Search API. Returns ranked search results with metadata, perfect for finding current information, URLs, and recent news.perplexity_ask
General-purpose conversational AI with real-time web search using thesonar-pro model. Great for quick questions and everyday searches with automatic citations.
perplexity_research
Deep, comprehensive research using thesonar-deep-research model. Ideal for thorough analysis, literature reviews, and detailed reports requiring multiple sources.
perplexity_reason
Advanced reasoning and problem-solving using thesonar-reasoning-pro model. Perfect for complex analytical tasks, logic problems, and chain-of-thought reasoning.
The
perplexity_reason and perplexity_research tools support an optional strip_thinking parameter. Set to true to remove <think>...</think> tags from responses, saving context tokens.Quick Links
Installation
Get started with installation and setup
Quick Start
Start using the MCP server in minutes
Tools Reference
Detailed documentation for all available tools
Requirements
- Node.js 18 or higher
- Perplexity API Key from the API Portal
- MCP-compatible client (Cursor, Claude Desktop, VS Code, etc.)