Source content inspection
Free LLM HTML visibility checker
Inspect the text and structure in a page’s initial HTML. Read the extracted sample, then compare it with the explanation visitors should see.
Use the LLM HTML visibility checker
Your input
No account required. Fair-use limits apply. This check analyses the input you provide.
Check the details. Then check the answers.
See when AI names your brand and which pages it cites in a free visibility report.
Get your free reportSee what is present in the source HTML
Choose the public page to compare
Enter the page URL and keep its browser version available so you can compare the content after the check.
Inspect the extracted source
Review a source-text sample, headings, content landmarks, structured-data syntax, script counts and indexing directives.
Investigate missing information
Compare the sample with the page you see in your browser. Check the source for important explanations that seem absent.
Scripts are not run. Source-text and script counts do not measure rendering dependence or AI-engine access.
Before the browser runs scripts
What does the original HTML actually contain?
A page can look complete in a browser while its initial response contains very little explanation. It can also contain plenty of source text that is mostly navigation. This checker lets you inspect the text and structural cues extracted from the HTML response itself.
The result includes word tokens, headings, main content landmarks, JSON-LD syntax, script counts, indexing directives and a sample of extracted text. It does not compare a browser screenshot with the source or reproduce the behavior of a particular LLM crawler.
What survives when the browser has not run your scripts?
Follow the page response into an extracted text sample, then compare it with the explanation a visitor should see.
The HTML returned for a public URL
Start with the response before browser scripts run. An empty application container gives the extractor little text to inspect.
A source specimen
Distinguish an application shell from an explanation.
Compare a minimal application shell with HTML that already contains a useful explanation.
Find the essential words in the HTML.
The checker extracts text outside excluded elements. If the main explanation is absent, investigate how it is delivered rather than assuming a browser view represents the initial response.
- Container<div id="app"></div>
- Script<script src="/app.js"></script>
- Review questionWhere does the product explanation enter the document?
Inspect the initial response before assuming the page’s main content is present.
A practical workflow
How to use the LLM HTML visibility checker
Submit the exact public page
Use a complete URL that returns HTML without authentication. A blocked endpoint or unsuccessful response cannot provide a reliable content sample.
Compare the sample with the page
Read the extracted words, headings and landmark findings. Open the page in a browser for your own comparison of the essential content.
Investigate delivery and access separately
Review rendering choices with a developer. Check robots policy separately, then test actual AI answers if visibility is the question you need to measure.
What extraction includes
Source text is not the same as rendered visibility.
The extractor omits script, style, template, noscript, textarea and head content. It also excludes explicit hidden, aria-hidden and inline-hidden elements. It does not resolve CSS classes, execute scripts or understand every malformed document in the same way as a browser.
Navigation and other boilerplate can remain in the extracted text. Use the sample and page context together. Word count is descriptive evidence, not a quality grade or a target you should inflate with repetitive copy.
Interpret scripts carefully
A script count cannot tell you what the scripts do.
The report observes script elements and the byte size of inline script content. Those figures do not measure execution cost, loading performance or actual rendering dependence. A page with many scripts may still contain all essential text in its HTML.
Conversely, a small application shell may need a substantial browser process to display the useful content. Investigate the rendered behavior directly. This checker’s result cannot prove that an AI service visited, understood or cited the page.
AI crawlability checker
Check robots rules for named crawler tokens alongside page directives and source content.
GEO audit
Review answer structure, source cues and publisher context in a public page’s HTML.
LLMs.txt generator
Build an editable llms.txt draft from one public page and selected links.
Questions, answered
LLM HTML visibility checker: your questions answered
A closer look at the inputs, method and next steps.
Does this show exactly what ChatGPT sees?
No. It shows what this extractor can read from a normal public HTML response. AI services may use different retrieval systems, cached content, access policies and rendering capabilities.
Does it compare source and rendered HTML?
No. Only the initial response is inspected. Open the page separately to compare what the browser displays after JavaScript, styles and other resources have loaded.
Why is my text missing from the sample?
It may appear after JavaScript runs, sit in an excluded element, occur later than the 1,600-character sample or be affected by the parser. Investigate the source and full page before choosing a fix.
Does a high word count mean good AI visibility?
No. The count can include menus and repeated boilerplate. It says nothing about relevance, source quality, crawler permission or whether an AI engine has used the page.
Are hidden elements always excluded correctly?
The extractor handles explicit hidden attributes and certain inline styles. It does not resolve stylesheet classes or browser layout, so some differences require a manual browser inspection.
Does this test robots.txt?
No. It observes page directives but does not fetch a robots policy for this tool. Use the AI crawlability checker when you need rule matching for named crawler tokens.
Can I use this for a private application?
The URL must be publicly accessible. The tool does not sign in or pass a visitor’s session, so private application content is outside the inspection.
Navigate the AI landscape. Place your brand on the map.
See when AI names your brand, which competitors appear, and the pages those answers cite. Turn what you find into a practical next step.
Five AI engines. No card required.