war.gov release
The U.S. Department of War's public UAP document release — the founding corpus of this archive, mirrored here with full text made searchable.
◆ About · Provenance & Method
A public record, kept open
UAP Archives is a free, searchable archive of declassified U.S. government documents on unidentified anomalous phenomena. 9,328 documents are indexed today, every one of them public domain. This page explains where the documents come from, how the archive is built, who maintains it, and how it stays online without tracking you.
What this is
For decades, the government's own UAP records sat in scattered FOIA reading rooms, agency PDFs, and archive boxes — technically public, practically unsearchable. This project pulls those releases into one place, runs them through text recognition and entity extraction, and puts a real search box, a map, a timeline, and an entity graph on top.
The goal is plain access. You should be able to find what a 1952 memo said about a sighting over a specific airbase without already knowing the file number.
Where the documents come from
NARA
The National Archives and Records Administration's UAP record series. We mirror metadata and link each match back to catalog.archives.gov, where the canonical file lives.
Project Blue Book
The U.S. Air Force's 1947–1969 UFO investigation case files, declassified and held at NARA — the largest single body of official sighting reports.
All of it is U.S. government work product and therefore public domain (effectively CC0) under 17 U.S.C. § 105. You can read, download, quote, and republish any document here for any purpose. We never add a license restriction the originals don't carry. The cross-source matching method is documented on the Sources page.
How the archive is built
- 01
Acquire
Documents come straight from the source: the war.gov UAP release, the National Archives (NARA) UAP record series, and the declassified Project Blue Book case files. Nothing here is leaked or scraped from behind a wall — it's already public record.
- 02
OCR
Most releases are scanned typewriter pages, carbon copies, and stamped memos. We run optical character recognition over every page so the text inside the image becomes searchable. OCR is imperfect on degraded scans — the original page image always stays one click away so you can check the machine against the paper.
- 03
Extract
A named-entity model reads the OCR text and pulls out the people, organizations, locations, and artifacts each document mentions. Those mentions are reconciled into canonical entities so 'Wright-Patterson AFB' and 'Wright Patterson' resolve to one place you can browse.
- 04
Connect
Dates, places, and entities let us cross-reference each document against outside sources — NARA, AARO, civilian sighting databases, and independent FOIA archives. The matcher and its scoring are documented in full on the Sources page.
- 05
Serve
Everything compiles to a static site on a global CDN — fast, no server to compromise, no account to create. Search runs in your browser. No tracking, no analytics that follow you.
Who maintains it
UAP Archives is built and maintained by an independent researcher. It is not affiliated with, endorsed by, or speaking for any government agency. Errors in OCR, entity extraction, or cross-source matching are ours, not the documents' — corrections and source suggestions are welcome.
How this is funded (and why there are no tracking ads)
Most free sites pay their bills by selling your attention: ad networks that profile you, trackers that follow you between pages, and a layout shaped to keep you scrolling. We made a deliberate choice not to do that. There are no ad networks here, no third-party trackers, and no analytics that build a profile of you. A site about government surveillance has no business running surveillance.
That choice has a cost. OCR over tens of thousands of scanned pages, entity extraction, search indexing, and bandwidth are real expenses. We cover them through direct support from people who value the archive — one-time contributions and optional memberships — not by monetizing your visit.
To be unambiguous: the documents are free and will stay free, forever. Support pays for the curation and convenience layered on top — never for access to the public record itself.
◆ Support the archive
Keep the record open
Every document here is public domain and stays free — no paywall, no login, no tracking ads, forever. What costs money is the work around the documents: OCR on tens of thousands of scanned pages, full-text search, entity extraction, and the bandwidth to serve it all to anyone who asks.
If the archive is useful to you, a contribution keeps it independent and surveillance-free. No pressure — the documents are yours either way.
- $5 Covers a day of OCR compute
- $25 A month of search hosting
- $100 A new release, ingested
Funds curation & convenience — never access; the documents stay free. Operated by Short Circuit LLC (for-profit), so support is not tax-deductible. See clearance tiers & what it funds →
◆ Dispatch
New releases, in your inbox
An occasional dispatch when a new tranche of documents is ingested or a notable case surfaces. Low volume, no tracking, unsubscribe any time.