Best SEO site checkers compared, on what each one actually reads

Site checker, site analyser, checkup tool and site auditor are four names for two genuinely different products, and knowing which one you are using decides whether the report is worth reading. One kind fetches your pages from your server, right now, and tells you what is in them. The other kind looks your domain up in a vendor's stored index and tells you what that index believes about you, which may be weeks old.

This page separates the two and then ranks inside each group on a stated criterion: what does the tool read, and does its own page publish the limit at which it stops reading? A tool that publishes its ceiling is one you can plan around. A tool that does not is one you discover the ceiling of halfway through an audit.

The two kinds, and how to tell which one you have opened

The site wide checkers, ranked on the published limit

Screaming Frog SEO Spider is first on this criterion because it is the clearest publisher in the category. Its own page states that the free version crawls up to 500 URLs in a single crawl, with no sign up or email capture, and names exactly what the annual licence adds: crawl configuration, saved and reopened crawls, JavaScript rendering, crawl comparison, near duplicate detection, custom extraction and the Search Console, Analytics and PageSpeed integrations. The same page counts the issue types it looks for at more than 300. A buyer can read that page and know in one minute whether free covers their site.

Search Console is second and it is first on a different criterion, which this page will not pretend otherwise about: it is the only tool that reports Google's own view rather than an inference about it. It is not a crawler of your whole site on demand, which is why the two sit side by side rather than one replacing the other.

The hosted checkers from the big suites come third for a small site and first for a large one. What you are buying there is scheduling and history: a crawl that runs on Tuesdays and tells you what changed since last Tuesday. That is a different product from a one off scan even when the issue list looks identical.

The single page analysers, and where they stop

A page analyser fetches one URL and reports the title, headings, word counts, internal and external links, images, structured data and speed. It is a fast and honest tool for one page. What it cannot do is compare the page with the pages ranking above it, which is the finding that decides whether the page moves, and almost every analyser presents its findings in its own scoring order rather than in the order that blocks the page.

The order that matters is fixed: indexability first, then whether the page works for a visitor, then the first paragraph and the title, then the content gaps against the top three, then internal links, then everything technical. Any analyser is useful if you read its output in that order and ignore its own.

The score out of 100, and why every one of them is different

Each tool computes a headline score by counting its own issues and applying its own weights. Two checkers will score the same site differently and both will change the number when the vendor ships a release. Google publishes no site score. Use the issue list, and judge progress on three real measurements instead: pages indexed against pages you intended to index, Core Web Vitals groups in the good band, and clicks on the pages you fixed.

The check that no site checker on this page performs

Run any checker above and you get a list of fixes for your own HTML. What you do not get is the answer a buyer sees when they ask an assistant to recommend somebody in your field, because that answer is composed for each asker and differs between apps, so there is nothing in your server for a crawler to find.

AI Knows Us is built for that single question. You supply the purchase questions; each one goes to ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, and the report is the list of businesses named in each reply. The unit of account is published on our pricing page: a credit buys one buyer question, on one app, asked once, with most checks spending a single credit, Claude spending three, and a Visibility API on the larger plans. We sell no crawler and no site checker, so on everything above this paragraph the picks in the tables are the answer and we are not a candidate. Ahrefs publishes a wider reading of the assistant surface than ours, including AI Mode and Copilot which we do not read, and publishes its own cost of 8 checks per update for a Claude check on its Brand Radar page.

Before paying anyone for a site check, it is worth knowing which ten checks you can run yourself in half an hour and how to read an agency's version of the same thing: free SEO analysis, the ten checks you can run today. On 21 October 2026 we counted durable inbound links across 1,655 tracked pages here and 696 of them had none. That page was one, which is part of why it is linked.

What buyers ask about SEO checkers and site analysers

Are free SEO checkers accurate?

For technical findings, yes, and the reason is that these are facts rather than judgements: a crawler either sees the broken link or it does not. The paid tiers differ on reach, on running to a schedule, and on holding data about sites you do not own, and the last of those is the only one you cannot replicate.

How often should I run a full site check?

Monthly, and immediately after a migration, a theme change, a template release or a platform upgrade. Those four events cause most sudden technical damage, and a check run the same day costs an hour instead of a quarter.

My checker reports thousands of issues. Where do I start?

Sort by the blocking order rather than the issue count. Most large numbers come from one template repeated across every page, so the real list is usually five or six fixes with a lot of multiplication behind them.

What is the difference between a site checker and a site audit service?

The tool produces findings; a service produces a prioritised plan and does the work. If you are buying a service, ask for the ticket list with URLs on each line, because that list is the deliverable and the PDF is the packaging.

Will a checker explain a drop in positions?

It will rule out the technical causes, which are the ones to eliminate first: a page fell out of the index, a canonical changed, a template started returning the wrong status, a redirect broke. Once the technical side is clean, the answer is a comparison with the pages now above you.