Lyceum Search

Lyceum Search

Zero API  ·  Zero Google  ·  76 US Cities  ·  9 Open Sources  ·  Security-filtered  ·  Top 20 by quality score

🔬 What this search engine does for you
🚫
Zero API Keys — Truly Independent

There are no API keys in this search engine at any level — not for you, not for the admin. Queries fire raw HTTP requests directly to open public indexes operated by non-profits, volunteers, and government agencies. No SerpAPI. No Google. No Bing. No search company involved in any step.

🌐
9 Open Sources — No Keys Needed

Common Crawl — non-profit, billions of pages.   Internet Archive CDX — Wayback Machine index, great for .gov & newspapers.   YaCy P2P — volunteer decentralised search network.   Data.gov — US federal open data (Census, EPA, BLS, HUD).   Wikipedia, Open Journals (DOAJ), Open Library, CrossRef, and Semantic Scholar — millions of academic papers, books, and research datasets.

🗺
City Tunnel — 76 US Cities

Results are geo-scored by how many city/state terms appear in their URL, title, and snippet. Only results mentioning your selected cities rank above zero — eliminating billions of irrelevant global pages and keeping results focused on a few million relevant US sites. Select individual cities, all metros, or all university towns at once.

🚫
Business Noise Excluded

Medical clinics, dental offices, auto repair shops, construction contractors, churches, home service companies, and known e-commerce spam domains are filtered out automatically. Over 60 keyword and domain patterns blocked at the pipeline level before any result appears.

📏
Content Depth Filter

Pages below your minimum word threshold are excluded. Word count is estimated from page byte size (~24 bytes per word of HTML). At the default 10,000-word minimum, you see only substantive content — no thin marketing pages, no stubs, no auto-generated filler.

🛡
Security & Spam Filters

Every URL is checked against malicious TLDs, phishing domain patterns, known exploit path signatures, and URL heuristics. Keyword spam (density >12%, low lexical diversity, title template abuse, slug stuffing) is detected and penalised. Results are sorted by a 0–100 quality score weighted by domain authority, source credibility, content depth, geo-relevance, keyword match, and freshness.