The web keeps falling apart. Pages go offline, accounts get deleted, papers slip behind paywalls, projects shut down. Usually a copy survives somewhere, in the Wayback Machine, archive.today, Common Crawl, scholarly indexes like Crossref, libraries like Open Library, and hundreds of smaller subject-specific archives. The catch is that they are all separate. To find out where something was saved, you used to open them one at a time: Wayback first, then archive.today, then a dozen academic and book databases.

Echo takes that manual round-trip off your hands. You paste one thing into a single box, a URL, a domain, a DOI, an ISBN, a handle, a file hash, or just a title, and Echo recognises what it is and hands you ready-made links into the archives that deal with that kind of thing.

One thing worth being clear about: Echo does not download or analyse anything for you. It does not open the pages or check whether a copy is really there. Its job is narrower and more useful than that, to work out what you pasted and show you which archives to go to for it. You follow the links yourself.

You never choose a category. Echo reads the shape of the input and picks the sources itself.

A page URL leads to that page's snapshots in the big web archives: the Wayback Machine (over a trillion pages), archive.today, Common Crawl, Perma.cc, Ghostarchive, and national archives. A bare domain opens the whole saved site rather than a single address, and adds a link to crt.sh, where a site's SSL-certificate history reveals its subdomains and its past. A DOI points to a paper in Crossref, OpenAlex and Semantic Scholar, and to open full text in Anna's Archive. An ISBN points to a book in Open Library, HathiTrust and Google Books. An @handle goes to the social sources, a file hash to Software Heritage and VirusTotal, an image link to reverse-image search.

No identifier? Type a title, a name, or a few words, and Echo fans the query out to every archive with a text search and sorts them by topic: books, papers, patents, images, museums, maps. "Frankenstein" goes to Open Library, Gutenberg and Wikisource; "attention is all you need" goes to arXiv, PubMed and DBLP.

How a search works

There is nothing clever about the flow. You paste, you press Search. Echo classifies the input and gathers every archive in its catalog that can take that kind of query. Then it groups them and gives each one a direct link, to the snapshot, the paper, the book, the profile. Above the results it shows what it took your input to be, so a wrong guess is easy to spot.

A result is not "here is your copy," it is "here is where to look for it." Echo walks you to the right door and you open it. Some archives are wide open, others hide their search behind a captcha or block robots outright; those still show up, with a direct link and a note on how to search them by hand. Where we can't take you straight to the result, we say so and point you to where it lives.

What not to expect

Echo holds no content of its own. The archives only keep what they managed to save before something vanished. If no public archive ever captured a page, a paper, or a video, there is nothing to pull back. Coverage is best for things that were public, popular, or cited.

The catalog

Alongside the search, Echo keeps a catalog of every archive we know about, including the ones we can't query automatically yet. You can browse it and filter by category, cluster, API, or access type, just to see what exists and what each archive holds.

Free, in six languages

Echo runs without JavaScript, so any search is a plain link you can save or send on. The interface comes in six languages: English, Russian, Spanish, German, Portuguese, French. The basics are free, and a free account raises your daily limit and keeps your search history.

Try it: echo.archivarix.net

The use of article materials is allowed only if the link to the source is posted: https://archivarix.com/en/blog/archivarix-echo/