Perplexity Search
Access the Perplexity search API for real-time information and answers.
Perplexity Search MCP Server
A Go implementation of a Perplexity Search MCP server that allows large language models (LLMs) to access the Perplexity search API through the Model Context Protocol (MCP).
Features
- perplexity_search: Perform web searches and return results, including citations
- Parameters:
query(string, required): The search querysearch_recency_filter(string, optional): Filter results by time (month,week,day,hour)max_tokens(integer, optional): Maximum number of tokens to returntemperature(number, optional, default: 0.2): Controls randomness in responsetop_p(number, optional, default: 0.9): Nucleus sampling thresholdsearch_domain_filter(array, optional): List of domains to limit search resultsreturn_images(boolean, optional): Include image links in resultsreturn_related_questions(boolean, optional): Include related questionstop_k(number, optional, default: 0): Number of tokens for top-k filteringstream(boolean, optional): Stream response incrementallypresence_penalty(number, optional, default: 0): Adjust likelihood of new topicsfrequency_penalty(number, optional, default: 1): Reduce repetitionweb_search_options(object, optional): Configuration options for web search
- Parameters:
Setup & Usage
Installing via Smithery
To install Perplexity Search Golang for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @chenxilol/perplexity-mcp-go --client claude
Prerequisites
- Go 1.23 or higher
- Perplexity API key
Installation
- Clone the repository:
git clone https://github.com/chenxilol/perplexity-mcp-go.git
cd perplexity-mcp-go
- Build the application:
go build -o perplexity-search-mcp
Running Locally
- Set your Perplexity API key:
export PERPLEXITY_API_KEY="your-api-key-here"
- Run the server:
./perplexity-search-mcp
Integrating with Claude
-
Copy the provided
claude_desktop_config.jsonto your Claude configuration directory:- Windows:
%USERPROFILE%\AppData\Roaming\Claude\claude_desktop_config.json - macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
- Windows:
-
Edit the configuration file to include your API key:
{
"mcpServers": {
"perplexity-search": {
"command": "/path/to/perplexity-search-mcp",
"env": {
"PERPLEXITY_API_KEY": "your-api-key-here"
}
}
}
}
Docker Support
- Build the Docker image:
docker build -t perplexity-search-mcp:latest .
- Run the container:
docker run -i --rm -e PERPLEXITY_API_KEY=your-api-key-here perplexity-search-mcp:latest
Example Usage
Once configured, Claude can use the perplexity_search tool via MCP to perform real-time web searches.
Example search with parameters:
{
"query": "latest AI research developments",
"search_recency_filter": "week",
"temperature": 0.5,
"return_related_questions": true,
"web_search_options": {
"search_context_size": "high"
}
}
Troubleshooting
If you encounter issues:
- Verify your API key is correctly set
- Check network connectivity
- Examine stderr logs for error messages
License
This project is licensed under the MIT License - see the LICENSE file for details.
Acknowledgments
- Model Context Protocol for the MCP specification
- MCP-Go for the Go MCP implementation
- Perplexity for their search API
Serveurs connexes
Scholarly
Search for academic articles from scholarly vendors.
RAG Documentation
Retrieve and process documentation using vector search to provide context for AI assistants.
Scientific Paper Harvester
Harvests scientific papers from arXiv and OpenAlex, providing real-time access to metadata and full text.
Enhanced PubMed Search
A search server for PubMed, the biomedical literature database, using a pure Node.js implementation.
MCP Research Friend
Research tools, including a Sqlite-backed document stash
Wikipedia
Retrieves information from Wikipedia to provide context to Large Language Models (LLMs).
Brave-Gemini Research MCP Server
Perform web searches with the Brave Search API and analyze research papers using Google's Gemini model.
Memvid
Encodes text data into videos that can be quickly looked up with semantic search.
vinted-mcp-server
First MCP server for Vinted marketplace. Search millions of second-hand items, compare prices across 19 countries (FR, DE, ES, IT, NL, BE, UK...), analyze seller profiles with ratings and verification status, detect cross-country arbitrage opportunities, and track trending products. 5 tools: search_items, get_item, get_seller, compare_prices, get_trending. 2 resources: countries list, categories. Works with Claude Desktop, Cursor, Windsurf, and any MCP-compatible client. Zero config — just npx.
MCP Web Search Server
A web search server powered by DuckDuckGo, no API key required.