Is it a real search engine or not? and why searxng ?
TL;DR: Only a handful of search engines actually crawl and index the web themselves (Google, Bing, Baidu, Yandex, Mojeek, Brave...). Most "alternative" engines quietly resell Bing's or Google's results. The notable exceptions: DuckDuckGo adds its own sources on top of Bing, and Ecosia + Qwant just launched Europe's first independent index in decades.
The basics: crawler vs. index
- A crawler (or spider) is a bot that visits web pages and follows links (Googlebot, Bingbot...).
- An index is the giant database of pages the crawler collects. That database is what a search engine actually searches.
Key insight: a search engine doesn't need its own index. It can simply license results from Google or Bing, or aggregate several engines (metasearch). Most engines you've heard of do exactly that.
Tier 1: Search engines with their own crawler and index
| Engine | Crawler | Notes |
|---|---|---|
| Googlebot | World's largest index | |
| Bing | Bingbot | 2nd largest index, and the most "resold" |
| Baidu | Baiduspider | Dominant in China |
| Yandex | YandexBot | Dominant in Russia |
| Naver | Yeti | South Korea |
| Daum (Kakao) | own | South Korea |
| Sogou | Sogou Web Spider | China |
| Qihoo 360 (so.com) | 360Spider | China |
| Shenma (Alibaba) | own | China, mobile-first |
| Seznam | SeznamBot | Czech Republic |
| Coc Coc | own + supplements | Vietnam, localized index |
| Mojeek | MojeekBot | Fully independent (UK), no Big Tech licensing |
| Brave Search | BraveBot / own API | 100% independent since late 2022 |
| Yep (Ahrefs) | YepBot | Growing independent index |
| Petal Search (Huawei) | PetalBot | Multilingual |
| Marginalia Search | own | Old-school, non-commercial web; open source |
| Wiby | own | Small, human-seeded web |
| Stract | own (started on Common Crawl) | Open source |
| Boardreader | own | Indexes forums only |
| Perplexity | PerplexityBot | Own index of hundreds of billions of pages |
| Wayback Machine | archive.org_bot | Historical archive, not live search |
Tier 2: Engines that run on Bing's index
This is the big one. Most "privacy-friendly" alternatives are actually Bing resellers:
- DuckDuckGo — Bing is its primary source, blended with its own DuckDuckBot, 400+ sources (Wikipedia, Apple Maps, WolframAlpha, Yelp...) and "Instant Answers". The dependency is real: a Bing API outage in 2024 took DDG down.
- Yahoo! Search — powered by Bing since 2009. New in 2026: Yahoo Scout, an AI answer engine built on Anthropic's Claude + Bing APIs.
- Ecosia — still Bing today, but transitioning (see below).
- Qwant — mostly Bing today, transitioning to the European index "Staan".
- AOL, Swisscows, Oscobo, Ekoru, KarmaSearch, Lycos — all Bing-powered.
Tier 3: Engines that run on Google's index
- Startpage — Google results, stripped of tracking (official partnership).
- Yahoo! Japan — Google technology.
- Gibiru — "uncensored" modified Google algorithms.
Tier 4: Metasearch / hybrid engines (no index of their own)
- MetaGer — aggregates mainly Bing, Mojeek, Brave.
- SearXNG — open-source metasearch (dedicated section below).
- Dogpile, Yippy — aggregators (Yippy clusters results from 30–40 engines).
- Kagi — hybrid: Google, Bing, Mojeek, Brave, Marginalia + its own indexing component.
- You.com — own indexing combined with AI and other sources.
- Presearch — decentralized, feeds off other engines' results.
- Openverse — indexes Common Crawl (a third-party open dataset), not a search engine.
AI search engines (2025–2026)
- Perplexity — built its own large-scale index and crawlers; no longer just Bing/Google.
- ChatGPT Search (OpenAI) — mostly Bing-based, with Google signals observed.
- Google Gemini / AI Overviews — Google's index.
- Microsoft Copilot — Bing's index.
What's new in 2025–2026
- EUSP / Staan — Ecosia (Germany) and Qwant (France) launched the first independent European search index. Live in France since August 2025 (30% of queries), Germany since July 2026. Staan (Search Trusted API Access Network) is open to other engines and AI companies.
- Perplexity went fully independent with its own index.
- Yahoo Scout (Jan 2026) — Claude + Bing APIs.
What is SearXNG?
SearXNG is a free, open-source metasearch engine — it doesn't crawl the web itself. Instead, it queries dozens of other engines (Google, Bing, DuckDuckGo, Mojeek, Brave, Wikipedia, Reddit, and 100+ more) through their public interfaces, merges the results, and serves them to you through a clean, ad-free page.
It's the community-maintained fork of the original SearX project.
Why self-host SearXNG?
| Advantage | What it means |
|---|---|
| Real privacy | Your queries never hit a third-party instance's logs. SearXNG is designed to not store your searches or IP addresses. |
| No middleman | Public SearXNG instances are run by strangers; their admins can technically see traffic. Self-hosting means the only person who sees your queries is you. |
| Avoid rate limits | Public instances get rate-limited and blocked by Google/Bing constantly. Your own instance on your own IP is far more reliable. |
| Full control | You choose exactly which engines to query, their ranking weights, which categories appear (images, videos, news, science...), and the UI theme. |
| Bypass regional filters | Pick engines by country/language and override local censorship or missing results. |
| No ads, no tracking | Self-hosted = zero advertising, zero analytics, zero profiling. |
| Browser integration | Set it as your default search engine in Firefox, Chrome, or Safari for private, unified search everywhere. |
| Learning + community | It's a well-documented Docker-friendly project with an active community and many public instances you can try before hosting your own. |
Who is self-hosting SearXNG for? Anyone privacy-conscious who wants Google-grade results without Google-grade surveillance. You don't need much: a small VPS, Raspberry Pi, or even your own computer with Docker is enough. Roughly 1 GB RAM will run it comfortably for personal use.
The honest trade-off: results are aggregated, so you get a mix of rankings from different engines rather than one polished ranking algorithm. Search quality is usually excellent for research-style queries, occasionally noisier than pure Google. For most people, that's a fair price for owning your own search.
Sources: SearchEngineWorld — Big Search Engine List (updated 30 Jul 2026), Ecosia blog — European search index, heise — Ecosia & Qwant launch German index, Wikipedia — DuckDuckGo, DITig — List of web search engines, SearXNG.