Extract URLs from text
Collect HTTP and HTTPS links and remove exact duplicates.
Your files are processed here, on your device
We do not send your file or text contents to a processing server. No account is needed and there is no daily operation quota. Individual tools have size limits to protect your device’s memory.
A connection is needed to load the site and its resources. Hosting may receive technical visit data. Favorites store only tool identifiers in this browser.
How we handle your data →How to use this tool
- Paste text containing complete HTTP or HTTPS links.
- Choose whether to keep or remove exact duplicates.
- Extract and review punctuation before copying or downloading.
An example you can verify
Collect links from text and inspect punctuation handling.
Example input
Visit https://example.com/a and (https://example.org/guia). https://example.com/a
Expected result
https://example.com/a https://example.org/guia
The closing bracket and sentence period are excluded from the second link. The repeated first URL appears once after deduplication.
Extract URLs from text: uses and practical guidance
Extracting URLs turns text containing scattered references into a list of links that can be reviewed separately. It helps inventory sources, collect addresses mentioned in notes or check which links appeared in a document. Matching recognizes addresses beginning with http:// or https://. A bare domain name is not assumed to be a link, and a protocol is not added automatically. This avoids turning words or incomplete names into addresses that the original text never supplied and keeps the operation focused on explicit references rather than guessing intended destinations.
Addresses retain their supplied spelling after common trailing punctuation is removed. Domains are not normalized, parameters are not reordered and query values are not decoded. Addresses that a server might treat as equivalent remain different strings when their spelling differs. Duplicate removal uses exact comparison and preserves the first occurrence. Keeping duplicates helps review repeated mentions, but does not establish how many actual links an HTML page contains. The tool examines pasted text instead of inspecting a hosted document or its link structure, so occurrences in source code and prose are both candidates.
Punctuation needs inspection. A period ending a sentence or a bracket surrounding an address usually belongs to the surrounding text, but those characters can also belong to a valid URL. Common trailing periods, commas and similar marks are removed along with surplus closing brackets, while balanced parentheses inside addresses are retained where possible. Automatic extraction cannot always determine the author's intent. If a final mark matters to a path or query, compare it with the source. Whitespace, quotation marks and HTML delimiters end a match. A link broken across lines is not automatically reconstructed into a continuous address.
Output contains syntactic candidates rather than availability checks. No link is opened, no server is queried, and redirects or destination content are not verified. Inspect protocols, domains and query data before sharing the list. An empty result can mean references lacked an explicit protocol. Input supports one million characters and ten thousand matches, with at most ten thousand characters per address. Retain source text when the context of each link matters. To inspect filters and values in one address, continue with the URL parameter parser after copying that URL, checking its original punctuation before interpreting its query.
Settings and limits
Paste text containing complete HTTP or HTTPS addresses. Common trailing punctuation and surplus closing brackets are removed. Keep or remove exact duplicates. Links are not opened and destination existence is not verified. Maximum 1 million characters, 10,000 matches and 10,000 characters per URL.
Frequently asked questions
Does it extract domains without a protocol?
No. example.com and www.example.com without http:// or https:// are not treated as complete links. Review matches when punctuation is part of the URL.