Link Extractor
Every Link, With Its Rel
This link extractor reads a page once and lists every link on it: where each goes, what it says, whether it leaves the site, and whether it carries nofollow. Then one click checks every link for a broken page or a redirect, and the whole list exports as CSV.
- FreePrice
- 5Filters
- CSVOutput
- NothingStored
Link Extractor
Every link on a page, with rel, destination and broken-link check
- Links found
- 0
- Internal
- 0
- External
- 0
- With nofollow
- 0
- Empty text
- 0
- Broken
- –
Asks every web link on the page for its status code, ten at a time. Only the reply headers are read, and redirects are reported rather than followed.
Links added by JavaScript after the page loads are not here, because this reads the page as the server sent it. That is also what a crawler reads.
Ready. Paste a public address and this server reads the page once.
What A Link List Tells You
Four numbers describe a page's linking: how many links, how many leave the site, how many carry nofollow, and how many have no anchor text at all.
Those four answer most of the questions people arrive with. For example: is this page linking out too much, are the outbound links marked, and can a screen reader read them.
The list underneath is where the detail sits. So each row carries the destination, the words, the rel value, and whether it opens a new tab.
Google's guidance on crawlable links is worth reading alongside it: a link only counts when it is a real anchor with a real href.
What This Reads, And What It Misses
It reads the page as the server sent it, which is also what a crawler reads. That makes the two lists comparable.
Links added by JavaScript after load are missing, and that absence is itself information. If your navigation is not in this list, it may not be in a crawler's either.
The anchor element is the only thing counted. A clickable div may act like a link, yet search engines never treat it as one.
Why Empty Anchor Text Matters
A link wrapped around an icon with no text announces itself as blank. Someone on a screen reader hears only the word "link".
Most pages have a few, usually in a header or a social row. However, the fix is an aria-label, or text hidden visually. Either takes a minute.
The filter here narrows to exactly those, which turns a vague accessibility worry into a list of five things to fix.
What The List Tells You About A Page
Count the internal links first. A page with two is stranded, and a page with ninety is a menu pretending to be an article.
Then read where they point. Good internal linking sends readers deeper, not sideways into the same three hub pages.
External links say something too. A page citing nothing at all usually asserts more than it can support.
And watch the anchor text column. Forty links reading here or this page mean the page fails for anybody skimming it.
Checking A Page You Do Not Own
Paste the address and the server fetches it, because your browser will not hand another site over to a script.
Only the extracted links come back, never the markup. So nothing from that page runs here.
Pages behind a login return whatever a signed-out visitor sees, which is usually a login form.
And a page built entirely by scripts may return very little. Where that happens, view the source, copy it, and use the paste box instead.
Internal Versus External
The tool splits links by their host name. Denwix treats a link to a subdomain as external. Most analysis treats it that way, and a subdomain is often a separate site.
Putting the CSV Export to Work
The CSV carries the current filter, so "external links with nofollow" comes out as its own file.
Feed it to a status checker to find the dead ones. Then compare two exports, before and after a redesign, to see what a template change removed.
For a competitor page, the external list shows who they cite, which is a faster way to find worthwhile sites than any directory.
Reading A Competitor Page
Their internal links show which pages they treat as important, since a page linked from every article is a page they are pushing.
Their external links show who they cite, which is a faster route to the worthwhile sites in a subject than any directory.
Their anchor text shows the phrases they want to rank for. That is stated plainly, in words they chose, on a page anybody can read.
All of this is fairly common. Reading what a public page links to is the oldest research method on the web.
One URL Each Time You Press Extract
This reads a single address rather than crawling a site. A crawler that follows links without permission is a different tool with different manners, and this is not one.
Pulling Every Link From a Web Page on Denwix
Three steps, and the filter is what turns three hundred rows into an answer.
- 1
Paste a public address
Any page a visitor could open. Pages behind a login cannot be read, because this server arrives as nobody.
- 2
Read the five counts
Total, internal, external, nofollow and empty. Those five describe the page's linking before you read a single row.
- 3
Narrow the list to the question in front of you
External only, nofollow only, or the ones with no anchor text. Each filter answers a different job.
- 4
Export what you filtered
The CSV follows the filter, so the file you download is the list you were looking at rather than everything.
What Sets This Link Extractor Online Apart
Plenty of tools list hrefs. The differences here are the rel values, the filters and what happens to the address afterwards.
rel values kept
nofollow, sponsored and ugc all survive into the list, so an audit of outbound links takes one look instead of thirty.
Anchor text included
The words matter as much as the destination, so an empty one counts as a finding in its own right.
Five filters
Internal, external, nofollow, empty text, or everything. After all, each one matches a question people actually arrive with.
Relative links resolved
Every href comes back as a full address, and therefore the list goes straight into a checker or an opener.
CSV that follows the filter
What you export is what you are looking at, which is rarer in these tools than it should be.
Nothing kept
The address serves one request and then goes. We keep no copy of the page and no record of the check.
Rules For Auditing Links On A Page
A list of links is raw material. These habits turn it into changes somebody can make.
- Start with the empty anchors. They make the smallest list and the clearest accessibility win, because every one of them is a five-minute fix.
- Check that outbound links are marked. Paid and affiliate links need sponsored. So filter to external, then read the rel column top to bottom.
- Run it before and after a redesign, then compare. Export both, then diff them. A template change that quietly dropped your footer links shows up in one line.
- Feed the export to a status checker. Dead outbound links are invisible until somebody clicks. The CSV is the input a checker wants.
- Look at the internal count too. A page with two internal links is a dead end, and most sites have obvious places to add three more.
- Look at the anchor text, not just the count. Thirty internal links all saying "here" is a worse finding than twenty-nine links, and it takes one glance at the column to see it.
- Read a competitor's external list. Who they cite tells you which sites in your subject are worth knowing, faster than any directory does.
- Watch for links a script adds. If your navigation is missing from this list, it is missing from a crawler's view as well. That is worth fixing.
- Review header and footer links on their own. Those links repeat on every page, so a fault there is a site-wide fault. The rest of the list is what makes this page different.
- Keep the CSV with the audit. A link list from six months ago is the only way to prove what changed, once somebody asks.
Six Findings People Misread
Each of these is easy to see in a list and easy to draw the wrong conclusion from.
Counting a missing link as a bug here
A link added by JavaScript is absent from this list and from a crawler's. A missing link is itself the result.
Treating every external link as a leak
Citing sources is how good pages work. Linking out is not a cost, since hoarding links has never helped a page rank.
Adding nofollow to everything
That was a fashion for years and it accomplished nothing. Use it where it means something, and leave ordinary links alone.
Ignoring the empty anchors
Icon links with no text are the most common accessibility failure on any page, though they are also the cheapest to fix.
Assuming a subdomain is internal
It is a different host, and most tools count it as external. If that matters, say which convention your report uses.
Crawling a site with a page tool
This reads one address. Running it a hundred times is not a crawl, and a real crawler asks permission first.
Extraction Questions People Ask
Will the Denwix link extractor read any public site?
Any public page. The checks refuse a private network, and a page behind a login stays out of reach, because this server arrives as an anonymous visitor.
Can it see links that scripts add after loading?
No, and neither do most crawlers. It reads the page as the server sent it, which is why a missing link here is worth investigating.
How many links can it read?
Up to a thousand from the first 512 KB of a page. The counts say when a page runs past that, so nothing is lost silently.
When does Denwix class a link as external?
Anything on a different host, including a subdomain of the same domain. That is how most link analysis treats it.
Can I save the link list as a CSV file?
Yes, as CSV, and the export follows whatever filter you have applied. Copy it or download it, whichever suits.
Why do some rows show no text?
Because the link wraps an image or an icon with no words in it. Filter to empty anchor text and you get exactly that list.
Will the extractor follow links across my entire site?
No. It reads one address at a time. A crawler is a different tool, and it needs to respect robots rules in a way a single fetch does not.
Does the address go anywhere else?
No. It serves one request and then goes, and this server writes no copy of the page anywhere.
What do the counts actually tell me?
Total says how heavy the page is on links. The internal and external split shows where it points, and the nofollow count shows how it treats what it points at.
Can I export only the external links?
Yes. Set the filter to external and export. The CSV follows whatever filter you are looking at rather than dumping everything.
Can I use this link extractor online for a competitor?
Yes. Reading what a public page links to is ordinary research, and the external filter is the fastest version of it.
Tools For A Link Audit
A list of links leads straight into these three.
HTTP Status Code Checker
Find out which of those links still work.
Open tool →Bulk URL Opener
Open the filtered list in tabs, in one press.
Open tool →Meta Tag Analyzer
Read the head block of the same page.
Open tool →
Ready to List Every Link on a Page?
Paste an address, filter to the question you have, and export the answer. The link extractor is free, needs no account, and keeps no record of the page.