Internal Link Checker Tools: What Each One Can Actually See

An internal link checker reads the links pointing from one page of your site to another and reports which work, which redirect, which break, and which pages nothing points to at all. The eight tools below all claim that job and do it at three very different scales. This is for an in-house SEO or content lead with fifty to nine hundred published URLs who has just been told the site has orphan pages. Every limit quoted below was taken from the vendor's own page, read on 5 October 2026.

The short version

• Orphan pages are found by subtracting one list from another, so it takes two inputs and never just one: a crawl, set beside an independent record of what you publish.
• These tools come in three scopes: one page, one crawl, or the whole graph. They are not substitutes for each other.
• A crawler that stopped at its page cap hands you orphans that are not orphans.
• Search Console shows you a sample of your links rather than all of them, and its tables stop at 1,000 rows.
• Four of the eight crawl a real website at no cost, and only one of those four asks for no account at all.

What does an internal link checker actually do?

Three different things, depending on how much of your site it is allowed to read. The category name hides that distinction, and the distinction is the whole decision.

Scope	What it can see	What it cannot answer
One page	Every link on the URL you are looking at, with status code, follow attribute and anchor text	Anything at all about a page you are not on
One crawl	The links between the pages the crawler actually reached	Pages the crawl never reached, for any reason
The whole graph	Every URL the tool knows you have, set against the ones something links to	Pages missing from both of its lists

Read the third row twice, because it settles most tooling arguments. An orphan page is defined by the links that are missing, not by anything on the page. A crawl only ever contains pages something linked to, so it cannot show an absence either. Finding an orphan means holding a crawl next to a second, independent record of what exists, usually an XML sitemap or Search Console, and subtracting.

How is this list ordered?

By three questions, in this order: how much of the link graph the tool can see, how much it reads before it stops, and how much friction stands between you and an answer.

Where those pull against each other, the tie goes to whatever reaches a true answer fastest for nothing. So a 500-URL crawler needing no account sits above a 5,000-page crawler that wants a verified domain. It also means two entries rank above GrowthHasten, because they read more of a website for free than its own free plan does.

One exclusion, stated rather than hidden: SEOptimer currently ranks first for this query and is not in the list. Its pricing page, homepage and internal link checker page each returned a bot-verification challenge on 5 October 2026 and did not render. Nothing is quoted from it because nothing was read, which is a statement about three attempts on one day rather than about the product.

Which internal link checker tool fits your site?

Tool	Scope	What you get free
1. Google Search Console	Whole graph, as Google recorded it	Everything, on a verified property
2. Screaming Frog SEO Spider	One crawl	500 URLs, no account
3. Hasten LinkLens	One page	Unlimited pages, one at a time, no account
4. Ahrefs Free	One crawl, up to its credit cap	5,000 Site Audit credits a month
5. GrowthHasten	Whole graph	A single site, 900-page ceiling, three re-analyses
6. Semrush Site Audit	One crawl	100 pages a month
7. Sitechecker	One page or one crawl	14-day trial only
8. Sitebulb	One crawl	14-day trial only

Each figure above and in the entries that follow was read off the vendor's page in a browser, rendered rather than fetched as raw markup, on 5 October 2026.

1. Google Search Console (Links report)

• Scope: The whole graph, as Google recorded it rather than as your HTML describes it.
• Pages or URLs read free: No page cap. Tables stop at 1,000 rows; the landing page exports up to 100,000.
• Account needed: Yes, and the property must be verified.
• What it reports: Top linked pages within your own site, and which pages link to a given page.

Google documents the gaps in this report more candidly than most vendors document theirs. Its Links report help page says the report is not comprehensive, that what you see is a sample, that URLs can drop out when they are unindexed or deduplicated, and that nofollow state is not shown. So treat it as the only record of what Google itself saw, and never as an inventory.

2. Screaming Frog SEO Spider (free version)

• Scope: One crawl, run from your own machine.
• Pages or URLs read free: 500 URLs. The paid licence is $279 per user per year.
• Account needed: None. Download it and run it.
• What it reports: Broken links, errors and redirects, plus titles, meta data, directives, hreflang and duplicates.

A real crawler rather than a demo, and 500 URLs covers a great many small sites outright. Screaming Frog's pricing page marks crawl configuration, saved crawls, JavaScript rendering and the Search Console and Analytics integrations as paid-version features. Its Orphan URLs filter is documented as URLs found only in an XML sitemap and not discovered during the crawl.

3. Hasten LinkLens

• Scope: One page, including pages behind a login or on staging.
• Pages or URLs read free: Any number of pages, one at a time, up to 2,000 links on each.
• Account needed: None. It is a free Chrome extension that checks the page you are on.
• What it reports: Five link states coloured in place, plus internal or external, dofollow or nofollow, page region and anchor text, with issues ranked.

The narrowest tool here, third because of one decision the others do not make: a 401, 403, 429 or bot-check response is reported as unable to verify, not as broken. A refused check and a dead link look identical in a status column and are completely different problems. The limit is the obvious one: it reads the page in front of you and nothing else.

4. Ahrefs Free (formerly Ahrefs Webmaster Tools)

• Scope: One crawl, which reaches the whole graph on any site inside its credit cap.
• Pages or URLs read free: Site Audit gets 5,000 crawl credits each month for every verified project, and you can verify any number of sites.
• Account needed: Yes, with verified ownership of the domain.
• What it reports: 170-plus technical and on-page issues, including broken internal and external links, and link type data for nofollow, UGC and sponsored.

No other free option here reads as many pages. Ahrefs also notes on its Webmaster Tools page that a credit is spent only when a crawled HTML page answers 200, which means broken URLs, redirects and non-HTML files never touch the allowance. Set against that, verifying the domain is a real step before anything runs.

5. GrowthHasten

• Scope: The whole graph. Its Links section is described as your internal link graph, and where a new link would help.
• Pages or URLs read free: A single website, 900 pages at most, and three re-analyses. No card and no trial clock.
• Account needed: Yes.
• What it reports: The link graph alongside 130-plus deterministic checks across 12 scored categories, plus sixteen months of Search Console history once that is connected.

Two entries above this one read more of a website for free: Search Console has no page cap, and Ahrefs Free reads 5,000 crawl credits a month against this free plan's 900. That is the honest reason GrowthHasten is fifth on a list ordered by how much you can see for nothing. What it does differently is a decision about what it refuses to report, two sections below. It does not track rankings and reports no search volume or keyword difficulty, so a blank in those columns is a real no rather than an oversight. Paid plans are coming. You can point it at a domain and read the link graph free today.

6. Semrush Site Audit (free account)

• Scope: One crawl.
• Pages or URLs read free: 100 pages a month.
• Account needed: Yes. An email and a password, with no credit card.
• What it reports: Technical issues including broken links and orphaned pages, with affected URLs and fixes.

Semrush's own guide to the free account says plainly that 100 pages will only cover a small portion of a large website, and suggests spending the allowance on the pages that matter most. For a link question that advice cuts badly: the pages most likely to be orphaned are exactly the ones you would not have chosen. A useful technical spot check, a weak link tool at this tier.

7. Sitechecker (Internal Link Checker)

• Scope: One page or one crawl. You choose per run.
• Pages or URLs read free: A 14-day trial. Its plans page lists Basic, Standard, Premium and Enterprise tiers and no free plan.
• Account needed: Yes. Email, Google or Facebook, with no credit card for the trial.
• What it reports: Status code, follow state and anchor text per link, plus internal backlinks to a given URL with the anchor distribution as percentages.

The anchor distribution view is the reason to look. Which phrases point at a page, and in what proportion, is a different question from whether the links work, and most of these tools skip it. If what the anchor text says is what you are investigating, this is the strongest reporting here.

8. Sitebulb

• Scope: One crawl, desktop or cloud.
• Pages or URLs read free: A 14-day trial. Its pricing page lists Lite at $18 a month for 10,000 URLs per audit, Pro at $42 for 500,000, and cloud plans from $125.
• Account needed: Yes, and a subscription once the trial ends.
• What it reports: Prioritised hints, crawl maps and duplicate content, with Analytics, Search Console and Sheets integration on every plan.

Last because its pricing page lists no free allowance of any kind, not because it reads less. The crawl maps are the clearest picture of a link graph here, and Search Console integration on the entry plan means proper orphan detection at $18 a month, which Screaming Frog's free tier cannot do at any price.

Which of these can find an orphan page, and which only looks like it can?

None of them can from a crawl alone. An orphan is a subtraction rather than an observation, and it needs two inputs: a crawl saying what is linked, and a separate record saying what exists.

• Both lists: Ahrefs Free, Sitebulb and Semrush each crawl and connect Search Console or Analytics, so they can compare the two. GrowthHasten crawls and takes Search Console as an optional connection.
• One list only: Screaming Frog's free tier crawls but cannot connect Search Console or Analytics, so its orphan route is the sitemap comparison, which finds only the orphans already in your sitemap.
• The second list but no crawl: Search Console records which pages link to a given page, so a page nothing links to has no row at all. The absence is the signal, and you go looking for it rather than being handed it.
• Cannot participate: a single-page checker, by definition. That is not a shortcoming in Hasten LinkLens or Sitechecker's page mode. It is what page scope means.

One practical consequence: if a tool reports zero orphan pages and you never gave it a sitemap or a Search Console connection, it did not check. It had nothing to check against.

Why do two tools report different internal link counts for the same page?

Five reasons, all design decisions rather than bugs, and together the most common reason a tool gets blamed for being wrong.

JavaScript rendering: a crawler reading raw HTML misses links a framework adds after load. Screaming Frog lists rendering as a paid-version feature, so its free tier and a rendering tool can legitimately disagree by a wide margin on one page.

Canonical and duplicate collapsing: the Links report groups pages by canonical URL and combines duplicate links from the same source to the same target. A crawler counting every <a> element separately returns a higher number, and neither is lying.

Nofollow handling: some tools count a rel="nofollow" link as internal and some exclude it. Search Console's documentation says its report does not specify whether a link is nofollow, so you cannot reconcile the gap from inside it.

Redirect chains: a link that passes through one or more redirects may be counted at its href, at its destination, or at both. Chains are where two counts drift furthest apart.

Refused checks reported as failures: a destination returning 401, 403, 429 or a bot challenge is not a broken link, but a checker that collapses everything non-200 into broken puts working links on your fix list.

What does a tool do when the crawl stops before the site does?

This is the failure mode nobody puts in a feature table, and the one that costs real time. Every crawler here has a ceiling: 500 URLs on Screaming Frog free, 5,000 credits a month on Ahrefs Free, 100 pages on free Semrush, 10,000 per audit on Sitebulb Lite. Over it, the crawl stops partway.

A truncated crawl does not produce a smaller orphan report. It produces a bigger and wrong one: a page reads as an orphan whenever the only thing pointing at it sat outside the crawl. An afternoon then goes on adding internal links to pages that already had them. So before trusting any orphan report, set the crawl's page count beside the number of URLs you believe you publish. If the first is smaller, you are holding candidates rather than findings.

GrowthHasten takes the other route and declines the job. Point it at a site larger than the free plan's 900 pages and the analysis is refused outright rather than run across part of the site, on the published reasoning that a score drawn from half a website still reads like a verdict on all of it. For link work that refusal is worth more than it sounds, because a declined crawl tells you nothing and a half-finished one tells you something false.

What will none of these tools do for you?

Write the link. Every tool here identifies, counts and ranks, and none of them edits your website; the ones that propose links propose them for a person to approve. GrowthHasten is explicit that it writes nothing to your website and that every suggestion is yours to judge.

• None of the eight knows which link is worth adding: a tool can say page A has one inbound link and page B has forty. Deciding A deserves more is a business judgement, and that judgement is most of what a real internal link strategy consists of.
• None of them reads intent: anchor relevance is scored on word overlap, not meaning, so a tidy anchor distribution can still describe the wrong page.
• Only one is a record of what Google did: only Search Console is, and even there it is a sample with a 1,000-row table limit.

Which one should you open first?

Start from the question rather than the tool.

• One page looks wrong, answer needed now: a single-page checker.
• Under 500 URLs, no budget: Screaming Frog's free tier, then the sitemap comparison for orphans.
• 500 to 5,000 URLs, no budget: Ahrefs Free, once you can verify the domain.
• The link graph joined to your Search Console figures: GrowthHasten's free plan under 900 pages, or Sitebulb if the site is larger and $18 a month is available.
• Only what Google recorded: the Links report, and nothing else answers it.

Before acting on any report from any of these, make the tool tell you what it reached. Set the crawler's URL count beside the figure you believe your site publishes, and treat a gap between the two as the first finding rather than a footnote. An orphan list built on an incomplete crawl is a list of guesses, and it reads exactly like a list of facts.