Kiwix Serve

Website: kiwix.org

kiwix-serve is the web server from the Kiwix toolset. It serves ZIM archives — compressed offline copies of Wikipedia, Wiktionary, Wikivoyage, Stack Exchange sites, Project Gutenberg and thousands of other collections — over plain HTTP, with a library page, full-text search and a top bar on every article. Download the ZIM files you want, point the server at them, and the whole lot is browsable on a LAN with no internet connection at all.

Key Features

• Serves one or many ZIM files: pass archive paths on the command line or maintain an XML library with kiwix-manage and serve it with --library; the library file can be reloaded automatically.
• Full-text search: searches the prebuilt index inside each ZIM, including multi-ZIM search.
• OPDS catalogue: the built-in catalogue endpoint exposes the library to OPDS clients, and --catalogOnly turns the instance into a pure catalogue server.
• Deployment options: custom port and URL root for a reverse-proxy subpath, thread count, per-IP connection limits, blocking of external links, and a custom welcome page.
• Companion tools: the same package ships kiwix-search for command-line full-text search and kiwix-manage for library maintenance.
• Offline by design: ideal for a home server, a school, a boat or anywhere the connection is missing or metered.

Licence

GPL-3.0 — free and open source. Note that the ZIM content itself is published by the Kiwix project under the licences of the underlying works; check the individual archive. If what you want is a library of your own e-books rather than offline copies of published collections, Calibre-Web is the closer fit.