Onion Search FAQs

Question 1

Can Onion Search be used with local LLMs and how is it deployed?

Accepted Answer

Yes, it's designed for local LLMs like those in LM Studio or Ollama. After a simple Python installation, it integrates as an MCP server with your LLM client. It operates a persistent session TUI (Text User Interface) that manages Tor connections and displays live activity.

Question 2

Does Onion Search support server deployments or automation?

Accepted Answer

Absolutely. Onion Search offers a headless mode, allowing it to run in the background without a visible TUI window. This makes it ideal for server deployments, automated LLM workflows, and integration into larger systems where a graphical interface isn't needed.

Question 3

What is Onion Search and how does it benefit LLMs?

Accepted Answer

Onion Search is a Model Context Protocol (MCP) server that empowers LLMs with anonymous web search and browsing capabilities through the Tor network. It ensures untraceable knowledge retrieval, providing your LLMs access to information without compromising privacy or exposing your identity.

Question 4

How does Onion Search guarantee anonymity?

Accepted Answer

It routes all web requests through the Tor network and utilizes a dedicated Tor Browser instance for page fetching. This makes your LLM's web traffic indistinguishable from millions of other Tor users, preventing browser fingerprinting, tracking, and geo-restrictions.

Question 5

What kind of web content can Onion Search retrieve?

Accepted Answer

It retrieves search results from DuckDuckGo, offering options to configure search region and safe search settings. For page content, it fetches and extracts cleaned text from specified URLs using Trafilatura via the Tor Browser, ensuring relevant and readable information for your LLM.

Onion Search

Onion Search

主な機能

ユースケース

主な機能

ユースケース