URL Extractor
Pull all links out of any block of text or HTML
0 unique URLs found
About URL Extractor
Scans a block of text or HTML and pulls out every web link it finds, useful for collecting all the links mentioned in a document, email thread, or page source without manually searching for and copying each one individually.
Paste your text or HTML, and every detected URL is listed automatically. Copy the list or download it as a CSV.
This catches links starting with http://, https://, ftp:// or www. — a bare domain such as "example.com" typed on its own isn't picked up, since it can't be reliably told apart from an ordinary word.
Watch: URL Extractor in Action
- Will this find a domain name that's typed without http:// in front of it?
- No — this catches full URLs starting with http:// or https://. A bare domain typed on its own in plain text (like just "example.com") isn't picked up, since there's no reliable way to distinguish that from a regular word ending in a domain-like suffix.
- Why was one of my links not listed?
- The most common reasons are that it has no https://, http://, ftp:// or www. start, that it is a mailto: or relative link, or that it is shorter than eight characters. A link glued directly to a preceding word or an at sign is also skipped.
- Which tracking parameters does Remove tracking take out?
- Any parameter starting with utm_, plus fbclid, gclid, msclkid, mc_cid, mc_eid, igshid, yclid and ref_src. Other query parameters and the part after # are kept. Links that become identical after cleaning are merged into one entry.
- What does the CSV contain?
- The first row is the header url,domain. Each following row has one entry and its domain, written without the protocol, path or www and in lower case. With Unique domains only switched on, both columns hold the host name. The file is called extracted-urls.csv.
- Can I paste HTML source instead of plain text?
- Yes. The pattern stops at spaces, angle brackets, double quotes and backticks, so addresses inside href and src attributes are picked up. Relative paths in those attributes are not, because they do not start with a protocol or www.
- Does it check whether the links work?
- No. It never connects to any of the addresses; it only recognises their shape in your text. A dead link and a live link are listed the same way, so test important ones in a browser or a link checker.
How to use URL Extractor
Paste any text, page source or exported document into the large box. Matching links appear straight away in the list under the buttons, and a counter above it reports how many unique URLs were found. Three switches change the list: Remove tracking strips campaign parameters and merges addresses that then become identical, Sort A-Z orders the list alphabetically, and Unique domains only replaces each link with just its host name, without https:// or www.
Press Copy All to place the list on your clipboard, one entry per line, or Download CSV to save a file named extracted-urls.csv with a url column and a domain column. Both buttons stay greyed out until at least one link has been found.
When to use it
Handy when you have a long email thread, set of meeting notes or newsletter and need every link in one place, for example to check them before publishing or to move them into a spreadsheet. It also works on pasted page source when you want to see which outside sites a page points to, and the domains switch turns that into a short list of sites.
When not to use it
It reads only the text you paste, so it will not crawl a website, follow links or test whether an address still works. Relative links such as /about and mailto: addresses are not collected, which makes it a poor fit for auditing a whole site's internal links. A site crawler or your browser's developer tools are better for that job.
Tips for best results
If you paste HTML, links inside href attributes are found as long as they start with http, https, ftp or www. Tracking removal runs first, then the domain reduction, then the sort, so turning on Remove tracking before exporting gives a tidy de-duplicated list for a report. A bare name like example.com on its own is ignored, but typing www. in front of it makes it count.
Limitations
Only addresses beginning with http://, https://, ftp:// or www. followed by a real-looking domain are matched; bare domains, mailto: links and relative paths are skipped. Duplicates are compared as exact text, so a link with and without a trailing slash counts twice. Trailing punctuation and an unmatched closing bracket are trimmed, which can occasionally clip a URL that genuinely ends that way. Matches shorter than eight characters are dropped.
Privacy
Matching is done by a pattern running inside your browser tab on the text you paste. Nothing is sent to a server to find or clean the links, and the tool never opens or visits any of them. The CSV is built in your browser and saved straight to your downloads folder. Close the tab and the pasted text and results are gone.
More from Text & Writing: