Does it only extract anchor tags?
No. Extract Links From HTML also surfaces image and media source URLs in addition to anchor links.
Can I export the result as JSON?
Yes. The extracted list is available as plain text and as structured JSON.
Does this upload my HTML anywhere?
No. Parsing happens entirely in your browser, so the markup you paste is never uploaded.
Is Extract Links From HTML free?
Yes. It is free to use, with no account and no upload required.
Can it handle a full page's source code?
Yes. You can paste large blocks of HTML, and it will extract the anchor, image, and media URLs it finds.
Does it extract iframe or stylesheet URLs?
No. The extractor queries a[href], img[src], and the src attribute on video, audio, and source elements. An <iframe src>, a <link rel="stylesheet" href>, and a <script src> are all outside that set and will not show up, even though they are URLs in your markup.
Why are some of my images missing from the list?
The selector is img[src], so an <img> that only carries srcset, or a lazy-loaded image that stores its real URL in data-src and leaves src empty, is not matched. That pattern is everywhere in modern responsive markup, so a page can genuinely have images and still return no image rows.
Are relative links turned into absolute URLs?
No. Each URL is copied out of the attribute exactly as written, so /pricing stays /pricing and ../img/a.png stays ../img/a.png. There is no page origin to resolve against, since the tool only ever sees the markup you pasted, and any <base href> in the source is ignored.
Why does the same URL appear twice?
Deduplication keys on the whole row, including the link text, rel, and target, not on the URL alone. Two anchors pointing at /about with different anchor text are treated as different rows. That is often what you want when auditing anchor text, and it does mean you should not read the row count as a unique-URL count.
Can it tell me which links are broken?
No. Checking a link means requesting it, and the tool makes no network requests at all. Everything here is derived from the markup in the box, so a URL that 404s is indistinguishable from one that works. Export the JSON and run the list through a link checker if you need status codes.
Is the HTML uploaded anywhere?
No. The markup is parsed in your browser and the extracted list is built locally, which is also why no link can be verified.