Empowers large language models with anonymous web search and content retrieval capabilities through the Tor network.
Sponsored
Many users leverage local Large Language Models (LLMs) for private and cost-effective information access. However, common LLM web integrations rely on standard browsers that leave identifiable fingerprints, compromising user privacy. This Model Context Protocol (MCP) server solves this by routing all web search and browsing requests through the Tor network and an actual Tor Browser instance. It ensures absolute anonymity by making user traffic indistinguishable from millions of other Tor users, enabling LLMs to access web information securely, privately, and free from tracking or surveillance.
Key Features
01Integrates with LLMs as a Model Context Protocol (MCP) server.
02Manages the Tor connection lifecycle automatically for optimal privacy and performance.
03Fetches web page content using a real Tor Browser instance for universal fingerprinting.
04Extracts and cleans text content from fetched web pages using Trafilatura.
05Performs anonymous web searches via DuckDuckGo through the Tor network.
060 GitHub stars
Use Cases
01Enabling LLM-powered applications to conduct web research while preserving user anonymity.
02Providing local LLMs with private and untraceable web access for information retrieval.
03Fetching specific web pages securely for LLM analysis without exposing user identity.
