tanigami/mcp-server-perplexity

Overview

This is a Python-based MCP server that provides a bridge for applications to interact with the Perplexity API via the Model Context Protocol. It manages API authentication and formats requests so that MCP-compatible clients can send queries and receive responses from Perplexity's language models.

Best for

Best for
Developers building MCP-based applications that need to leverage Perplexity's AI for real-time queries or research.

Use cases

• Integrating Perplexity AI into MCP-compatible chat or research tools
• Building a custom assistant that queries Perplexity for up-to-date information
• Automating API calls from applications that support the Model Context Protocol

How to use

Tools exposed

• ask_perplexity

Tested with

Claude Desktop

Example client config

{\n  "Perplexity": {\n    "command": "uvx",\n    "args": [\n      "mcp-server-perplexity"\n    ],\n    "env": {\n      "PERPLEXITY_API_KEY": "your-perplexity-api-key"\n    }\n  }\n}

Notes

This is a Python-based MCP server that provides a bridge for applications to interact with the Perplexity API via the Model Context Protocol. It manages API authentication and formats requests so that MCP-compatible clients can send queries and receive responses from Perplexity’s language models.

92 stars on GitHub. Last updated 2024-12-25. Licensed MIT.

Use cases

• Integrating Perplexity AI into MCP-compatible chat or research tools
• Building a custom assistant that queries Perplexity for up-to-date information
• Automating API calls from applications that support the Model Context Protocol

Pros

• Straightforward setup for any MCP-compatible environment
• Gives access to Perplexity’s search-enhanced AI through a standard protocol
• Open-source with a small but active community (92 stars)

Cons

• Requires a Perplexity API key and incurs usage costs
• Dependent on network connectivity and Perplexity API uptime
• Limited to the capabilities and knowledge cutoff of Perplexity’s models

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.