SearXNG — Open Source Dark Web Metasearch Engine Guide (2026)

What Is SearXNG?

SearXNG is an open-source metasearch engine — instead of building and maintaining its own index, it queries multiple other search engines simultaneously and combines the results into a single unified response. A single SearXNG query might aggregate results from DuckDuckGo, Bing, Google, Wikipedia, GitHub and dozens of other sources depending on how the instance is configured — returning a more comprehensive result set than any single engine provides.

It is self-hostable — anyone with a server can run their own SearXNG instance. Several community operators run instances as .onion hidden services, providing access to SearXNG’s aggregation capabilities within the Tor network with no central operator, no persistent logging and fully auditable open-source code.

There is no single “official” SearXNG .onion address. The service is distributed across many independent instances run by different operators with different configurations. Finding a good current instance is part of using SearXNG effectively.

Finding Current .onion Instances

How to find current .onion instances:

• Visit searx.space from any browser — it lists all publicly known SearXNG instances with status, uptime and configuration details
• Filter by “Tor” or “.onion” in the instance list to find Tor-accessible instances
• Search “searxng onion” on Ahmia — active .onion instances appear in results
• Check Dread’s /d/Onions subforum for community-recommended current instances

Why there’s no single address: SearXNG’s decentralized architecture means any address published here would be for one specific instance operated by one specific party. That instance may change configuration, go offline or be taken over at any time. The instance directory at searx.space is the authoritative source for current, actively maintained instances.

How SearXNG Works

When you submit a query to a SearXNG instance, the instance simultaneously queries multiple configured search engines on your behalf and aggregates the results. From your perspective it looks like a single search — from the network’s perspective, multiple queries are going out to multiple sources.

How to Use SearXNG

1. Open Tor Browser with security level set to Safer — most SearXNG instances use JavaScript for result rendering
2. Find a current .onion instance via searx.space or Ahmia
3. Navigate to the instance .onion address
4. Enter your search query
5. Select search categories — General, Images, News, Files, Social Media etc. depending on instance configuration
6. Review aggregated results from multiple sources on a single page

SearXNG Instance Quality — What to Look For

Not all SearXNG instances are equal. Instance quality varies based on configuration, maintenance and operator reliability. Several indicators help identify well-maintained instances:

SearXNG vs. Single Dark Web Search Engines

Self-Hosting SearXNG as a .onion

Advanced users can run their own SearXNG instance as a personal .onion hidden service. This provides the strongest available privacy — you are the instance operator, so no third party has access to your queries even at the instance level.

Running a personal SearXNG instance requires a server or VPS, Docker or manual installation, Tor daemon configuration to expose it as a .onion service and ongoing maintenance for updates. The Tor Project’s documentation covers exposing services as .onion addresses — SearXNG’s own documentation covers the installation process.

For most users, using a well-maintained community instance is practical enough. Personal instance hosting makes sense for security researchers, journalists or organizations that handle sufficiently sensitive queries to justify the operational overhead.

Frequently Asked Questions

Which SearXNG instance should I use?

Check searx.space for current instance status — filter for .onion instances and select one with high uptime, many configured engines and a recent SearXNG version. Do not rely on any single instance long-term — instance operators change configuration, retire instances or go offline without notice. Bookmark two or three good instances as backups.

Does SearXNG find .onion sites?

It depends on the instance configuration. SearXNG can be configured to query dark web search engines like Ahmia alongside clearnet engines — instances configured this way will return .onion results. Instances configured for clearnet sources only will not return .onion results. Check the instance’s configured engines list before assuming .onion coverage.

Can I trust a SearXNG instance not to log my queries?

SearXNG’s open-source code does not log queries by default — but instance operators can modify the configuration. The code is auditable but the running instance is not. For the strongest privacy, use a .onion instance run by a party you have reason to trust, or run your own instance. Accessing any instance via the .onion address protects your IP from the instance operator — it does not protect query content from a logging instance operator.

Is SearXNG the same as SearX?

SearXNG is a fork of the original SearX project. The original SearX development stalled and SearXNG emerged as an actively maintained continuation with improvements to the interface, engine support and privacy features. Most current instances run SearXNG rather than the original SearX. The two are compatible but SearXNG is the recommended current version.

Why use SearXNG instead of just using multiple search engines separately?

Two reasons. First, convenience — a single query returns results from multiple sources simultaneously rather than requiring separate queries on each engine. Second, privacy — each search engine you query directly receives your query even if your IP is hidden through Tor. SearXNG’s instance IP replaces yours for all upstream queries, meaning individual search engines cannot build a profile based on your specific query patterns even from anonymous connections.