How to use
- Paste the text, HTML or Markdown that contains the links.
- Press the button to extract them.
- Read the list: one URL per line, in the order they appeared.
- If you expected more results, check whether the missing links start with www. but no http:// (see below).
- Copy the list into a spreadsheet or link checker.
- Remove duplicates afterwards if the same link appears several times.
Formula / how it works
Step 1 — find candidates. The tool searches for the pattern https?://[^\s<>"']+ (case-insensitive). In words: "http://" or "https://", followed by every character up to the next space, angle bracket or quote. Stopping at < > " and ' is what lets it pull links cleanly out of HTML such as <a href="https://site.pk/page">.
Step 2 — trim sentence punctuation. Links in prose are often followed by a full stop, comma, colon, semicolon, question mark or exclamation mark that belongs to the sentence, not the address. Those characters are removed from the end of each match. "see https://example.com/a?b=1, and" yields https://example.com/a?b=1. A question mark inside the URL, before a query string, is kept, because only trailing characters are trimmed.
Step 3 — balance brackets. A closing bracket at the end is removed only if the URL has more closing than opening brackets. That keeps Wikipedia-style addresses such as /wiki/Lahore_(city) intact while dropping the bracket when a whole link was written inside parentheses.
What is not extracted:
- Bare domains like www.google.com or toolsfreehub.com without a scheme. Detecting these reliably needs a list of top-level domains and produces many false positives (file names like report.pdf, abbreviations like e.g.).
- Other schemes such as ftp://, mailto: and tel:. Use Extract Emails for addresses.
- Links split across lines by an email client's line wrapping; only the first part is found.
- Relative links in HTML, such as href="/about", because they have no scheme.
Duplicates and order. Every occurrence is listed, in reading order. If a newsletter repeats the same tracking link ten times, you get ten lines. Paste the result into Remove Duplicate Lines if you need a unique list.
Encoded and tracking links. Many emails wrap destinations in redirect URLs (for example a link-tracking domain followed by an encoded target). The tool returns the wrapper as written; decoding the real destination needs a URL decoder and sometimes a visit to the link.
Validity is not checked. The tool finds text that looks like a URL. It does not contact the site, so a link can be listed even if it is broken, expired or malicious. Check suspicious links with a link expander or safe-browsing service before clicking.
Worked examples
Links followed by sentence punctuation
Inputs: Docs at https://example.com/a?b=1, and http://test.org/x.
Paste the sentence and press the button. The comma after the first link and the full stop after the second are trimmed; the ? inside the first URL stays because it is not at the end.
Result: https://example.com/a?b=1
http://test.org/x
Older versions kept the comma and full stop.
HTML with a bare domain
Inputs: See <a href="https://site.pk/page">here</a> or www.google.com
Paste the HTML. The match stops at the closing double quote of the href, so the tag is not included. www.google.com has no scheme, so it is not extracted.
Result: https://site.pk/page
Add https:// to bare domains if you need them.
Brackets that belong to the URL
Inputs: (https://en.wikipedia.org/wiki/Lahore_(city)) end
Paste the text. The raw match ends with "))". It has one opening bracket and two closing ones, so one closing bracket is trimmed, leaving the balanced Wikipedia path.
Result: https://en.wikipedia.org/wiki/Lahore_(city)
The outer bracket belonged to the sentence.
Use cases
- Collecting reference links from a research document or chat.
- Auditing outbound links in HTML before publishing.
- Pulling download links from release notes or emails.
- Building a list of pages for a link checker or crawler.
- Reviewing suspicious links in a message without clicking them.
Tips & limitations
- Add https:// to bare domains before extracting if you need them.
- Use Remove Duplicate Lines on the output for unique links.
- Rejoin links that an email client wrapped across lines.
- Do not click unknown links; check them with a safe-browsing tool.
- Paste HTML source rather than rendered text to catch href links.
FAQ
Why was www.example.com not found?
Only links starting with http:// or https:// are extracted. Bare domains are skipped to avoid false matches like file names.
Does it remove trailing full stops and commas?
Yes. Sentence punctuation at the end of a link is trimmed; punctuation inside the URL is kept.
What about URLs with brackets?
A trailing closing bracket is removed only if it is unbalanced, so Wikipedia-style links keep their brackets.
Are duplicate links removed?
No. Each occurrence is listed in order. Remove duplicates afterwards if needed.
Can it extract mailto or ftp links?
No, only http and https. Use Extract Emails for email addresses.
Does it check whether links work?
No. It only finds link-shaped text; nothing is fetched.
Runs entirely in your browser; your text is not uploaded. Counts and scores are heuristics, so review the output before you publish it.