- Home
- MCP servers
- Perplexity
Perplexity
- javascript
284
GitHub Stars
javascript
Language
6 months ago
First Indexed
2 months ago
Catalog Refreshed
Documentation & install
Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.
Installation
Add the following to your MCP client configuration file.
Configuration
View docs{
"mcpServers": {
"dainfernalcoder-perplexity-mcp": {
"command": "node",
"args": [
"/path/to/perplexity-server/build/index.js"
],
"env": {
"PERPLEXITY_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}You can run the Perplexity MCP Server to access an intelligent research assistant that routes tasks to the most suitable model and provides search capabilities for every task. This lets you perform quick lookups, complex analyses, and in-depth research from a single MCP integration.
How to use
To use this MCP server, connect it to an MCP client and start one of the provided server configurations. You can run the server locally with Node.js or via NPX, and you’ll supply your Perplexity API key for authentication.
// Option 1: Run with a local node process (explicit build index)
{
"mcpServers": {
"perplexity": {
"command": "node",
"args": ["/path/to/perplexity-server/build/index.js"],
"env": {
"PERPLEXITY_API_KEY": "YOUR_API_KEY_HERE"
},
"disabled": false,
"autoApprove": []
}
}
}
// Option 2: Use NPX to run a prebuilt MCP package
{
"mcpServers": {
"perplexity": {
"command": "npx",
"args": [
"-y",
"perplexity-mcp"
],
"env": {
"PERPLEXITY_API_KEY": "your_api_key"
}
}
}
}
// Option 3: Pass API key via command-line, then start the server
{
"mcpServers": {
"perplexity": {
"command": "node",
"args": [
"/path/to/perplexity-server/build/index.js",
"--api-key",
"your_api_key_here"
],
"disabled": false,
"autoApprove": []
}
}
}
// Option 4: Use a .env file from a project directory with --cwd
{
"mcpServers": {
"perplexity": {
"command": "node",
"args": [
"/path/to/perplexity-server/build/index.js",
"--cwd",
"/path/to/your/project"
],
"disabled": false,
"autoApprove": []
}
}
}
How to install
Follow these concrete steps to set up the Perplexity MCP Server locally. Ensure you have Node.js installed and an API key for Perplexity. Then clone the project, install dependencies, and start the server using one of the provided methods.
# Prerequisites
- Install Node.js from nodejs.org
- Obtain a Perplexity API key from your Perplexity account
# Install and prepare
- git clone <repository-url>
- cd <cloned-directory>
- npm install
- npm run build
# Start using the stdio method (example 1)
# Ensure you replace the path with your actual build location and API key
node /path/to/perplexity-server/build/index.js --api-key YOUR_API_KEY_HERE
# Start using NPX (example 2)
npx -y perplexity-mcp
# Start using the explicit internal path (example 3)
node /path/to/perplexity-server/build/index.js --cwd /path/to/your/project
Configuration notes and security
All configurations require the Perplexity API key to be provided either via an environment variable, a command-line argument, or a .env file. The priority order is command-line argument, then environment variable, then a .env file provided with --cwd.
Environment variable to set: PERPLEXITY_API_KEY. Use a strong key and keep it secret. If you are using a local project directory with a .env file, place the key in that file and ensure your tool picks it up.
Tools and capabilities
The server exposes three capabilities to handle different task types with Perplexity models.
-
searchQuick lookups and concise answers for straightforward questions. -
reasonComplex, multi-step reasoning for explanations and comparisons. -
deep_researchIn-depth research and detailed reports on major topics.
Troubleshooting
If the MCP client cannot parse the API key automatically, supply it directly via the command-line or ensure the environment variable PERPLEXITY_API_KEY is set correctly in your runtime environment.
Available tools
search
Quick search tool for simple queries and concise information lookup.
reason
Handles complex, multi-step tasks requiring detailed analysis, explanations, and comparisons.
deep_research
Conducts comprehensive research and generates detailed reports for in-depth topics.