AI SEO Audit

What is the AI SEO Audit?

The AI SEO Audit is a free tool that measures sixteen SEO signals on one page, such as HTTPS, title, meta description, canonical, JSON-LD and image alt text, then has an AI model grade those measurements. You get an overall score and letter grade, four sub-scores, a passed or failed note for each check, and up to five fixes.

How to use the AI SEO Audit

1. Enter the URL of the page to audit, such as https://example.com/blog/post. The audit reads that exact page.
2. Press Audit website. The tool fetches the page, times the fetch, and checks whether /robots.txt, /sitemap.xml and /llms.txt exist on the site.
3. Press Sign up free to unlock to see the result. The audit has already run; a free account reveals it.
4. Read the overall score and grade, then the four scores: Crawler Speed, Technical SEO, Meta Tags and Content Structure.
5. Go through the sixteen checks, each marked passed or failed with one sentence on why, and then the recommendations.

What does an AI SEO audit check?

This one checks sixteen things on a single page. The tool measures them in code: HTML fetch time, HTTPS, whether /robots.txt, /sitemap.xml and /llms.txt exist, JSON-LD, the lang attribute, the title, the meta description, the canonical, Open Graph and Twitter card tags, a favicon, the H1 count, how many images have alt text, and how many internal links the page has.

Those values then go to an AI model, which marks each check passed or failed, explains it in a sentence, and sets the scores and grade. The measuring is done in code so the numbers are real. The model only judges what it is given.

Why doesn't the AI visit the page itself?

Because a chat model asked to audit a URL will usually produce an audit anyway. It will name your title tag, estimate your load time and comment on your headings for a page it never fetched. The output is fluent, specific and invented, and it is worse than no audit because it reads exactly like a real one.

So the prompt here carries raw values and nothing else. If the result says your meta description is 68 characters, that is because the tool counted 68 characters.

How are the AI SEO audit scores worked out?

The overall score, the letter grade and the four sub-scores are set by the AI model from the measured values, under rules the tool writes into the prompt. Crawler Speed is tied to how long the HTML took to fetch: 100 under 500 milliseconds and 50 at 2,000 milliseconds. A missing robots.txt or sitemap lowers Technical SEO, a missing or badly sized title or meta description lowers Meta Tags, and the H1 count and alt-text coverage decide Content Structure.

Each sub-score ring is labelled Excellent at 80 or more, Good from 50 to 79 and Needs Work below 50, and the overall rating is coloured on the same bands. Because a model applies the rules, two runs on an unchanged page may not give exactly the same numbers. The pass or fail on each check rests on a measured value, so that is the steadier thing to compare between runs.

How is an AI SEO audit different from a traditional SEO audit?

A traditional SEO audit tool crawls a whole site, following links from page to page and reporting things like status codes, redirect chains and duplicate titles across hundreds of URLs. This audit reads one page in depth and adds the signals that matter for AI search alongside the classic ones: whether an llms.txt file exists, whether the page carries JSON-LD, and whether it declares a language.

The "AI" part is also in who writes the result. The checks are measured in code, then a model turns the raw values into plain sentences and a ranked list of fixes, so you get a reason next to every fail instead of a bare red cross. It does not replace a site crawl. It tells you, quickly, what is wrong with one page and in what order to fix it.

What does this AI SEO audit not check?

It reads the HTML the server sends and does not run JavaScript, so copy, links or tags added in the browser are invisible to it, as they are to most AI crawlers. robots.txt is fetched and its rules are read for eight major AI crawlers, including GPTBot, ChatGPT-User, ClaudeBot and PerplexityBot, and the AI Crawlers Allowed result comes from those rules. /sitemap.xml and /llms.txt are checked only for existence at the root of the domain, so a sitemap kept at another path shows as missing. To test robots.txt rules against a specific page path, bot by bot, use the AI Bot Crawlability Checker.

Speed here means the time to fetch the HTML, not how fast the page renders for a visitor; the Website Speed Test covers load time. Internal links are counted, not tested, so use the Broken Link Checker to find dead ones. Two counting rules are worth knowing: an image with an empty alt="" counts as having no alt text, and only links whose href starts with / or with your full domain count as internal.

How do I fix what the AI SEO audit finds?

Fix the failures that stop a crawler first, then the ones that shape how the page is described, then structure. In practice that means HTTPS, robots.txt and the sitemap before the title, meta description and canonical, and those before Open Graph tags, alt text and internal links. The recommendations list, up to five items, names the specific flaws found on this page, so it is a reasonable starting order.

Run the audit on one page of each template, such as an article, a product page and a category page, because a fault in a template repeats on every page built from it. After you deploy a fix, run the audit again on the same URL and check that the failed item now passes.