URL Extractor
Links hide inside articles, chat logs, emails and documents — surrounded by prose, sometimes repeated, often with trailing punctuation stuck to them.
Paste an article, a document or a page of HTML and this URL extractor finds every link inside it — useful for link audits, research and checking which URLs a page actually points to.
URLs Found
How It Works
The extractor scans your text for anything starting with http:// or https:// and reads until it hits a space, a line break, or a character that cannot appear in a URL (such as a quote, an angle bracket, or a closing parenthesis that was not opened inside the URL). It then trims trailing punctuation — a full stop, comma, or semicolon at the end of a sentence is not part of the link, so "see https://example.com." gives you https://example.com without the dot.
Two clean-up steps run after extraction. Remove duplicates drops repeated URLs — the comparison is exact, because two URLs that differ by a single character can point to different pages. Sort alphabetically orders the list so duplicates and near-duplicates sit next to each other and the list is easy to scan. The result table reports how many URLs were found in total, how many were unique, and how many duplicates were removed.
The output is one URL per line, ready to paste into a spreadsheet, a link checker, a bookmark import, or a document. If a URL contains spaces or special characters, a URL encoder makes it safe to share. Extraction runs entirely in your browser — the text is never uploaded and no link is ever fetched or checked. Use it to audit the links in a draft, collect sources from research notes, or pull every link out of a long thread. The same approach works for email addresses — an email extractor pulls every address out of a block of text. The same copy-and-paste approach works for browser data — a user-agent parser breaks a UA string into browser, version, operating system and device.
Worked Example
The example text contains https://www.example.com/guide twice, https://www.example.com/faq once, and http://blog.example.org/post-1 once — 4 matches in total, of which 3 are unique. Sorted alphabetically, the list is http://blog.example.org/post-1, https://www.example.com/faq, https://www.example.com/guide. Note that the trailing full stop after the last URL is not included in the result.
How to Use This Tool
Paste any text and click Extract URLs. The links appear one per line, with the counts underneath.
- Paste your text. Anything works — an article, a chat log, an email, a document, or a copied web page.
- Choose the clean-up options. Deduplication and sorting are both on by default. Switch deduplication off if you want to see every occurrence, including repeats.
- Copy the list. Copy All copies the URLs one per line. Paste them into a spreadsheet, a link checker, or a bookmark manager.
The example on the page contains four mentions of three URLs, so you can see deduplication working. Reset restores it.
Tips and Practical Advice
Extraction is the easy part — a few details decide whether the list is actually useful.
- Only http(s) links are found. Bare domains without a scheme ("example.com") are not extracted, because plain text is full of dotted words that are not links. Paste text that includes the http:// or https:// prefix.
- Trailing punctuation is trimmed. A period, comma, semicolon, or closing parenthesis at the end of a URL is treated as sentence punctuation, not part of the link — so links at the end of sentences come out clean.
- Extraction is not a link check. A well-formed URL can still be broken, redirected, or gone. This tool pulls links out of text; it never fetches them or checks whether they work.
- Use the unique count for audits. When checking how many distinct links a page or document contains, the unique count — not the total — is the number that matters.
What Counts as a URL? (and What Gets Left Out)
A URL has five parts, and knowing them explains what this tool keeps and what it drops. Take https://www.example.com/guide?page=2#top: the scheme (https://) says how to fetch it; the host (www.example.com) names the server; the path (/guide) names the page; the query (?page=2) passes extra parameters, usually for search results, filters or tracking; and the fragment (#top) points to a section within the page. Change any one of those parts and you have a different URL — which is why this tool compares URLs exactly, character for character, when it removes duplicates.
That exactness matters in practice. A link with a tracking parameter — https://example.com/a?ref=home — is a different URL from the same link without it, even though both land on the same page. The same goes for fragments: https://example.com/a#top and https://example.com/a point to the same page but are different strings, so both survive deduplication. Worked example: a text that mentions https://example.com/a twice, https://example.com/b once, http://example.org/x once, https://example.com/a?ref=home once and https://example.com/a#top once contains six mentions but five unique URLs — only the exact repeat of https://example.com/a is removed. If you want those near-duplicates merged, strip the query strings and fragments in a spreadsheet first, then dedupe.
Two things the extractor deliberately does not do. First, it does not find bare domains: "example.com" on its own is not extracted, because plain text is full of dotted words that are not links — version numbers like 2.4.1, file names like report.pdf, and abbreviations. Requiring the http:// or https:// prefix keeps false positives out. Second, it never fetches a link: extraction is a text operation, not a link check. A well-formed URL can still be broken, redirected or gone, and finding that out needs a link checker that actually requests each page.
Extraction is most useful as the first step in a link audit. Pull every URL out of a draft before publishing to check for typos and duplicates; pull the links out of research notes to build a source list; pull them out of a long thread or chat log to save the references before they scroll away. Sorting the output groups links by domain, so a quick scan shows which sites a document leans on most.
Sources: No external data is used on this page — all processing runs in your browser on the text you enter. Last reviewed: 21 November 2026.
Frequently Asked Questions
How do I extract URLs from text?
Paste the text into the box and click Extract URLs. Every link starting with http:// or https:// is pulled out, deduplicated, and listed one per line with a count — copy the list with the Copy All button.
Does it remove duplicate URLs?
Yes — deduplication is on by default, using an exact comparison: two URLs that differ by even one character count as different links, because they can point to different pages. The result table shows how many duplicates were removed.
Why is a trailing full stop not part of the URL?
Links at the end of a sentence usually pick up the sentence's punctuation. The extractor trims trailing periods, commas, semicolons, and closing brackets so the copied link works when pasted into a browser.
Does it check whether the links work?
No. This tool only pulls URLs out of text — it never fetches them or checks their status. A well-formed URL can still be broken or redirected; use a link checker for that.
Can I extract URLs from a PDF or Word document?
The tool works on text you paste, not on files. Copy the text out of the document first — select all, copy, paste here — and the extractor will pull out every http:// or https:// link it contains.
Can I extract just the domains, without the full URLs?
No — the tool returns complete URLs, because the path and query are part of the link. Sort the output alphabetically to group links by domain, or paste the list into a spreadsheet and split on the first slash after the domain.
Last updated: