SearXNG

privacy

Privacy-respecting metasearch engine that aggregates results from 70+ sources without tracking queries or storing personal data. Self-host for a private search experience or use a public instance

#search#privacy#meta-search#self-hosted
Alternative to GoogleBingDuckDuckGo

Quick Start

docker run -d -p 8080:8080 -v searxng:/etc/searxng searxng/searxng:latest

Overview

SearXNG is a metasearch engine that sends your query to multiple search engines simultaneously, merges the results, and returns them with no tracking, no query logging, and no advertising profile built from your searches. It is a fork of the original SearX project, with more active development and a noticeably cleaner default interface.

The core mechanism is query proxying. When you search, SearXNG forwards the request to Google, Bing, DuckDuckGo, and whichever other engines you have enabled. From the perspective of those engines, the request comes from your server, not from you personally. Your IP address, search history, and browsing habits are not correlated across searches because SearXNG does not store any of it.

You can configure which engines contribute to each result category. Image searches, news searches, and general web searches each have their own engine pool. Specialist engines for code, science papers, maps, and video are also available and can be enabled per query with special syntax.

The result quality limitation is honest and worth stating plainly. SearXNG does not improve on the underlying engines — it aggregates them. If Google is returning poor results for a query, SearXNG will include those poor results alongside Bing’s. The value is privacy and breadth, not better search algorithms. For most everyday queries the aggregated results are comparable to using a single engine directly.

Public SearXNG instances are available at searx.space if you want to try it without hosting. For anyone who searches frequently enough that privacy matters, a self-hosted instance on a small VPS is a ten-minute setup.

SearXNG: Pros & Cons

Pros (The Wins)Cons (The Friction)
70+ engine sources:
Broader coverage than any
single search engine alone.
No ranking intelligence:
Result quality depends on
upstream engines, not SearXNG.
Zero tracking:
No query logs, no profile,
no advertising correlation.
Rate limiting:
Google blocks high-volume
instances; CAPTCHAs appear.
Configurable engines:
Weight or disable sources
per category or query.
No personalisation:
No learned preferences;
same results for everyone.
30.8k stars:
Active SearX fork with
cleaner UI and faster releases.
Public instances degrade:
Shared instances often
rate-limited by Google.

Use Cases

Specific ways to use SearXNG for your workflow.

01
Search the web without Google or Bing tracking your queries and building an advertising profile on you
02
Run a shared search instance for a team or household on a private server with no query logging
03
Aggregate results from multiple engines to get broader coverage than any single search engine provides
04
Use as the default search engine in your browser with no personalisation or filter bubble effects

Deployment Strategy

Recommended ways to host SearXNG in your own environment.

docker
self-hosted