Self-hosted link archive with tags and monitoring
Free
Tool review — Self-hosted · Updated

Self-hosted web archiving in HTML, PDF, and WARC
Free
ArchiveBox is an open-source, MIT-licensed web archiving tool you run on your own server or machine. Instead of sending pages to a cloud service, it fetches URLs you feed it — one at a time, from browser bookmark or history exports, RSS feeds, Pocket, Pinboard, or plain text files — and saves each one in multiple durable formats: raw and single-file HTML, PDF, PNG screenshot, WARC, extracted article text, and media downloaded via yt-dlp. Under the hood it drives headless Chromium, wget, SingleFile, and Readability, storing everything in an SQLite database plus plain files on disk. It is a mature, community-maintained project with over 28,000 GitHub stars, offering a CLI, self-hosted web UI, Python API, and a browser extension.
Your own server (self-hosted archive folder) · SQLite database + on-disk snapshot files · Optional remote storage backends (S3, NFS, SMB)
ArchiveBox is completely free and MIT-licensed, with no official paid tier — you supply your own server and storage. If you don't want to run it yourself, third-party managed hosting providers such as PikaPods and StellarHosted offer ArchiveBox instances starting around $3/month.
Self-hosters, researchers, and archivists who want permanent, local copies of web pages in open formats rather than links stored in a third-party cloud service. Comfort with Docker or the command line is expected.
Yes. ArchiveBox is open source under the MIT license with no paid tier; your only costs are your own server and storage. Third-party managed hosting is available from around $3/month if you don't want to self-host.
Each URL is saved in multiple formats: raw HTML, self-contained single-file HTML, PDF, PNG screenshot, WARC, extracted article text, favicons, headers, and media files downloaded via yt-dlp.
Native Windows installation is not supported. Windows users typically run ArchiveBox through Docker; Linux and macOS are supported directly via Docker Compose, pip/uv, Homebrew, or apt.
Yes. It imports URLs from browser bookmark and history exports, RSS feeds, Pocket, Pinboard, JSON/CSV files, and plain text lists, and can run scheduled imports automatically.
Something it wouldn’t save, or a format it mangled? Tell us what happened →