Markdown Link Extractor (All Links, Images, References)
Pull every link out of a Markdown document — inline [text](url), reference-style [text][ref], autolinks <url>, and bare URLs. Useful for link audits, broken-link checks, importing references into another tool.
How to use the Markdown Link Extractor (All Links, Images, References)
Paste Markdown. The tool extracts all link references and renders as list / CSV / JSON.
Every link form Markdown allows
Markdown can express a link four different ways: the inline [text](url) form, the reference form [text][label] with the URL defined elsewhere, an autolink <https://example.com>, and a bare URL that renderers turn into a link on their own. A long document mixes all of them freely, which makes it hard to see at a glance every destination it points to.
This extractor scans the whole document, finds links in all four forms, resolves reference labels to their actual URLs, and lists each link with its anchor text. You can include images in the output or leave them out, and export the result as a plain list, CSV, or JSON — handy for auditing destinations, feeding a broken-link checker, or importing references into another tool.
Common use cases
- Link audits — list every destination a document points to in one pass.
- Broken-link checks — export the URLs and run them through a checker to find dead links.
- Content migration — collect all links before moving a post to a new platform.
- Building reference lists — pull the citations or resources out of a long write-up.
- Separating internal and external — review the URL list to see which point off-site.
Frequently asked questions
Which link forms does it detect?
[text](url), reference [text][label], autolinks <url>, and bare URLs. Reference links are resolved to the URL defined in their [label]: url line.How is this different from the image extractor?
[text](url) form. The Markdown image extractor targets image references, the  form. You can also include images here with the toggle if you want both.Can I export as CSV or JSON?
Does it resolve reference-style links?
[text][label] link is matched to its [label]: url definition elsewhere in the document so the listed URL is the real destination, not the label.