URL to anything
Any web page, in whatever shape you need.
Shorten a link, then add an ending: .pdf for a printable copy, .md for the text, .jpg for a full-page screenshot, .json for the facts, and + for an archive page with citations. One link, every format, and a saved copy that stays when the page changes.
What you get
URL to anything
The page as a document
A PDF as it prints, a full-page screenshot as JPG, PNG or WebP, a reading view, a web archive, all from one load of the page in a real browser.
The page as data
Markdown and plain text for notes and assistants, JSON with the page's title, words, facts and citations, its tables as CSV.
A copy that stays
The saved copy keeps what the page said the day you saved it. When the page changes or goes away, your link still has something to show.
Citations ready
The archive page (+) writes the APA, MLA, Chicago and BibTeX citation, pointing at the saved copy, with the date it was taken.
Versions over time
Save the page again when it changes and keep each version: 10 a link on Free, 100 on Pro, each one fingerprinted.
Fast, and shared
Creating the link answers at once; the copy fills in seconds. A page anyone saved in the last hour comes back instantly and costs nothing.
How it works
Three steps, no setup.
- 01
Shorten the page on the home page with Save switched on, at /archive, or from a terminal. You get the short link and its QR code.
- 02
Add an ending to the link for the format you need: .pdf, .md, .txt, .jpg, .json, .csv, or + for the archive page.
- 03
Share it or cite it with the citation from the archive page. Save the page again later and both versions stay, side by side.
# Shorten a page; the first .md request saves it curl -d url=https://example.com/report https://snip.link curl https://snip.link/<code>.md # The same page as a PDF, a screenshot, and data curl -O https://snip.link/<code>.pdf curl -O https://snip.link/<code>.jpg curl https://snip.link/<code>.json # The archive page: versions, citations, fingerprints open https://snip.link/<code>+
The same from the REST API and the MCP server, inside Claude, ChatGPT, Cursor or Claude Code.
Free to start. Pro for volume.
Everything here works on the free plan. Pro is $4 a month or $36 a year, for more room and longer history.
| What | No account | Free | Pro |
|---|---|---|---|
| Pages saved a day | 10 | 50 a day | 1,000 a day |
| Versions kept per link | 5 | 10 a link | 100 a link |
| Every format by its ending | Yes | Yes | Yes |
| Citations and fingerprints | Yes | Yes | Yes |
| Private copies, only you can open | No | Yes | Yes |
| Watch a page and email what changed | No | No | Up to 50 pages |
Who it is for
Built for the way you share.
Research and journalism
Cite a source with a saved, dated, fingerprinted copy. The quote is still there when the page is edited.
AI assistants and agents
Hand an assistant a link ending in .md and it reads the page, without a scraper, a key or a browser of its own.
Developers
Any public page as Markdown or JSON with one GET, from a URL you can also hand to a person.
Questions
About pages in any format.
Which endings work on a link?
.md, .txt and .html for the text, .pdf for a printable copy, .jpg, .png and .webp for a full-page screenshot, .json and .csv for the data, .mhtml for a web archive, .svg and .qr for the QR code, .rss for the versions, and + for the archive page. Add ?download to save any of them as a file.
Is it free?
Yes. 10 saved pages a day without an account, 50 on a free account and 1,000 on Pro. Every format and every citation is included on every plan.
What does the saved copy look like?
The page as a real browser rendered it, in every format at once: Markdown, plain text, a reading view, a full-page screenshot, a PDF and a web archive. Each file carries its SHA-256 so anyone can check it is unchanged.
Does it work on pages behind a login or a paywall?
No. The copy is what a visitor with no account sees. Pages that ask not to be archived are respected and skipped.
Can I keep a saved page private?
Yes, on an account: a private copy opens for you only, while the short link still redirects for everyone else.
How is this different from a web archive site?
The copy lives at a short link you own, with a QR code, in every format, with a citation written for you, and it is made the moment you ask through the site, the API or an assistant. There is no queue and no separate address to remember.