Perplexity MCP Zerver
@wysh3
About Perplexity MCP Zerver
MCP web search using perplexity without any API KEYS
Config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"perplexity-mcp-zerver": {
"command": "bun",
"args": [
"run",
"build"
]
}
}
}Tools
No tools detected
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
Overview
What is Perplexity MCP Zerver?
A minimalist research server implementing the Model Context Protocol (MCP) to deliver AI-powered research capabilities through Perplexity’s web interface. It provides keyless operation via browser automation, making it suitable for developers who want to search, extract content, and analyze code without traditional API keys.
How to use Perplexity MCP Zerver?
Install the Bun runtime and Node.js 18+, then clone the repository, run bun install and bun run build. Configure your MCP client by adding a server entry with "command": "bun" and "args": ["/absolute/path/to/build/main.js"]. Initiate research by sending commands through your MCP client, such as “Use perplexity to research quantum computing advancements.”
Key features of Perplexity MCP Zerver
- Intelligent web search and summarization without API limits
- Persistent conversation history stored in local SQLite
- Clean article extraction with native GitHub repository parsing
- Developer tooling: documentation retrieval, API discovery, code analysis
- Keyless operation using browser automation
- Optional Perplexity Pro login for better models and higher limits
Use cases of Perplexity MCP Zerver
- Researching topics like quantum computing advancements via natural language queries
- Retrieving technical documentation for frameworks such as React 18
- Discovering relevant APIs for development needs
- Analyzing code snippets for deprecated patterns
- Maintaining context-rich, persistent conversations on technical subjects
FAQ from Perplexity MCP Zerver
How does authentication work?
No API key is required; the server uses browser automation to interact with Perplexity’s web interface. An optional Pro account login can be performed to unlock better models and higher limits.
What are the runtime prerequisites?
Bun runtime and Node.js 18+ (for TypeScript compilation) are required. The server is built with TypeScript and runs under Bun.
How is data privacy handled?
All processing is local: research results are stored in a local SQLite database, and browser profile data resides in the directory specified by PERPLEXITY_BROWSER_DATA_DIR.
What does the optional Pro account provide?
Logging in with a Perplexity Pro subscription grants access to GPT-5.1, Claude Sonnet 4.5, and other enhanced models, as well as higher usage limits.
Are there any limitations?
Stability depends on Perplexity’s website consistency. The project is intended for educational use only, and browser automation may break if the site changes.
Frequently asked questions
How does authentication work?
No API key is required; the server uses browser automation to interact with Perplexity’s web interface. An optional Pro account login can be performed to unlock better models and higher limits.
What are the runtime prerequisites?
Bun runtime and Node.js 18+ (for TypeScript compilation) are required. The server is built with TypeScript and runs under Bun.
How is data privacy handled?
All processing is local: research results are stored in a local SQLite database, and browser profile data resides in the directory specified by `PERPLEXITY_BROWSER_DATA_DIR`.
What does the optional Pro account provide?
Logging in with a Perplexity Pro subscription grants access to GPT-5.1, Claude Sonnet 4.5, and other enhanced models, as well as higher usage limits.
Are there any limitations?
Stability depends on Perplexity’s website consistency. The project is intended for educational use only, and browser automation may break if the site changes.
Basic information
More Search MCP servers
G-Search MCP
jae-jaeA powerful MCP server for Google search that enables parallel searching with multiple keywords simultaneously.
🚀 OneSearch MCP Server: Web Search & Crawl & Scraper & Extract
yokingma🚀 OneSearch MCP Server: Web Search & Scraper & Extract, Support agent-browser, SearXNG, Tavily, DuckDuckGo, Bing, etc.
Serper Search and Scrape MCP Server
marcopesaniSerper MCP Server supporting search and webpage scraping
Kagi MCP server
kagisearchThe Official Model Context Protocol (MCP) server for Kagi Search & other tools.
Version 2.0 is here!!
mixelpixxMCP Server built for use with Claude Code, Claude Desktop, VS Code, Cline - enable google search and ability to follow links and research websites
Comments