shahar42/Perplexity_custom_mcp_server

Perplexity MCP Server

A lightweight MCP (Model Context Protocol) server that integrates Perplexity AI with your development environment. Ask real-time questions and get AI-powered answers with live web search and citations.

Features

• 🌐 Real-time Web Search - Get up-to-date information with citations
• 🤖 Multiple Models - Choose from 5 different Perplexity models
• 💬 Easy Integration - Works with Claude Code and other MCP clients
• ⚡ Fast & Lightweight - Minimal dependencies, quick setup
• 🔒 Secure - Your API key stays local

Available Models

Quick Start (One-Line Installation)

Prerequisites

• Linux or macOS
• Python 3.8+
• curl (usually pre-installed)

Installation

1. Clone the repository

git clone https://github.com/yourusername/perplexity-mcp-server.git
cd perplexity-mcp-server

2. Run the installation script

chmod +x install.sh
./install.sh

3. Enter your Perplexity API key when prompted
   • Get your free API key at https://www.perplexity.ai/

That's it! ✅

Manual Setup (If needed)

If the script doesn't work for you:

# Create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Setup environment
cp .env.example .env
# Edit .env and add your API key
nano .env

Usage

Start the Server

source venv/bin/activate
python server.py

With Claude Code

Add to your Claude settings (typically ~/.claude/settings.json):

{
  "mcpServers": {
    "perplexity": {
      "command": "path/to/perplexity-mcp-server/venv/bin/python",
      "args": ["path/to/perplexity-mcp-server/server.py"],
      "env": {
        "PERPLEXITY_API_KEY": "your-api-key-here"
      }
    }
  }
}

Then restart Claude and you'll have access to:

• perplexity_ask - Ask questions with configurable models and parameters
• list_models - See all available models

Example Queries

See for detailed examples including:

• Basic queries and information requests
• Advanced reasoning and step-by-step analysis
• Long-form research and synthesis
• Custom parameter usage
• Real-world workflows

Quick examples:

Basic Question

Ask Perplexity: "What are the latest developments in AI (2025)?"

Advanced Reasoning

Ask with sonar-reasoning model: "Explain how transformers work in machine learning"

Research

Ask with sonar-deep-research model: "Write a comprehensive summary of renewable energy trends"

API Key Setup

1. Go to https://www.perplexity.ai/
2. Sign up or log in
3. Get your API key from the API settings
4. Run ./install.sh or add it to .env:

   PERPLEXITY_API_KEY=your_key_here
   

Troubleshooting

"PERPLEXITY_API_KEY environment variable required"

• Make sure .env file exists and contains your API key
• Run: cat .env | grep PERPLEXITY_API_KEY

"Curl failed" or "HTTP 401"

• Your API key is invalid or expired
• Get a new one from https://www.perplexity.ai/

Python version issues

• Make sure you're using Python 3.8+
• Check with: python3 --version

venv activation issues

On some Linux systems:

source venv/bin/activate
# or
. venv/bin/activate

How It Works

The server uses:

• FastMCP - For the MCP protocol implementation
• curl - For direct HTTP requests to Perplexity API
• python-dotenv - For secure environment variable management

No heavy dependencies - just what you need!

Contributing

Found a bug? Have a suggestion? Please open an issue on GitHub!

License

MIT License - feel free to use and modify

Questions?

Check the Perplexity API docs for more details on models and parameters.

Made for developers. By developers.