Perplexity Search

Perplexity Search is an MCP server that connects large language models to the Perplexity search API using a lightweight Go implementation. It is designed for developers, researchers, and automated assistants needing access to current web results and citations directly within client workflows. By bridging the Model Context Protocol and Perplexity, the server exposes the perplexity_search tool, allowing models to query the live web with granular control. Users can configure parameters such as recency filters by hour, day, week, or month, restrict searches to specific domain lists, control token limits, and adjust sampling temperature or repetition penalties. It also supports streaming responses, returning image links, and retrieving related search questions. Because it is compiled as a standalone binary in Go or executed inside a Docker container, it offers fast execution and minimal resource overhead, making it suitable for running locally alongside desktop LLM environments or within containerized agent architectures.

Category: Web Search & Research

Tags: perplexity, real-time, search, web search

Visit Perplexity Search

How to install and configure Perplexity Search

1. Ensure you have Go 1.23 or higher installed, and obtain a Perplexity API key. 2. Clone the repository and build the binary: bash git clone https://github.com/chenxilol/perplexity-mcp-go.git cd perplexity-mcp-go go build -o perplexity-search-mcp Alternatively, install automatically via Smithery for Claude Desktop: bash npx -y @smithery/cli install @chenxilol/perplexity-mcp-go --client claude 3. Add the server to your claude_desktop_config.json file: json { "mcpServers": { "perplexity-search": { "command": "/path/to/perplexity-search-mcp", "env": { "PERPLEXITY_API_KEY": "your-api-key-here" } } } } 4. Restart Claude Desktop to enable the Perplexity Search tool.

What you can do with Perplexity Search

• Conducting time-sensitive web searches by applying recency filters to capture breaking news or research updates from the past day or week. * Restricting search outputs to verified documentation or specific websites using the domain filter parameter to improve research accuracy. * Gathering citations, source links, and relevant reference materials alongside synthesized answers during deep analytical tasks. * Fetching related follow-up questions and image URLs directly into conversational AI interactions for expanded topic exploration.

Key facts

• https://github.com/chenxilol/perplexity-mcp-go
• Web Search & Research
• perplexity, real-time, search, web search

Part of MCP Servers

How do I install Perplexity Search?

You can install Perplexity Search by cloning the repository and compiling the binary with Go 1.23 or higher, running it through Docker, or using the Smithery CLI installer to automatically configure it for Claude Desktop. You will need a valid Perplexity API key set in your environment or client configuration file.

What can Perplexity Search do?

Perplexity Search provides the perplexity_search tool for performing web searches. It supports parameters to control token outputs, temperature, top-p, top-k, recency filtering across hours, days, weeks, or months, domain-specific filtering, streaming responses, and retrieving source citations, related questions, and image links.

Which MCP clients work with Perplexity Search?

Perplexity Search works with any client that adheres to the Model Context Protocol, including Claude Desktop. Configuration instructions and example configuration files are provided for Claude Desktop on Windows, macOS, and Linux.

Is Perplexity Search open source?

Yes, Perplexity Search is open-source software licensed under the MIT License. The complete Go source code and Docker setup files are publicly available on GitHub at https://github.com/chenxilol/perplexity-mcp-go.