FREE CHECK / NO SIGN-UP

A TOOL BY SHARPFONT / DESIGN & AI STUDIO

Can AI read
your site?

More people now ask ChatGPT, Claude or Perplexity before they search. Those tools can only repeat what their crawlers were able to read. Enter your address and see exactly what they receive from your page.

  • Free, no sign-up
  • Result in seconds
  • Your address is not stored

A FREE TOOL BY SHARPFONT · CLIENTS SINCE 2010

WHY IT HAPPENS

People see the finished page. A crawler sees the code.

Two pages can look the same in a browser and be very different in what they send.

IN YOUR BROWSERAFTER THE SCRIPT HAS RUN

The page looks complete. A person never notices that the text was added after loading.

WHAT AN AI CRAWLER RECEIVES0 WORDS
<body>
  <div id="root"></div>
  <script src="/main.8f2a.js"></script>
</body>

The text only appears after the script has run. A crawler that does not run it receives an empty page.

A static page sends every word in the code itself, so browsers, phones and crawlers all get the same page.

WHY IT MATTERS

Three ways a good site goes unread.

  1. Your page looks finished. To a crawler it is blank.

    Sites built as JavaScript apps often send an almost empty page and fill it in afterwards. You see the finished page. Most AI crawlers only get the empty one.

  2. robots.txt says welcome. The firewall says no.

    Hosts and firewalls can turn AI crawlers away at the server, before your site is reached. Nothing in your own files shows it. One line in robots.txt can do the same in plain sight.

  3. You cannot see either problem. This check shows both.

    It reads your page the way a crawler does, asks for it under each crawler’s name, and tells you which of the eight checks fail.

WHAT WE LOOK AT

Eight checks. Three groups.

Each one is read from the page code, your robots.txt and your server’s answers: the same things a crawler meets.

01 / TEXT

The text

How many words reach a crawler that does not run JavaScript, and whether the headline, title and description are in the page code.

02 / ACCESS

The door

Whether your robots.txt shuts out the crawlers of OpenAI, Anthropic, Perplexity and Google Gemini, whether your server turns them away, and whether the page is marked “do not index”.

03 / STRUCTURE

The signposts

Structured data, image descriptions and real links, which help a machine understand the page and find the rest of your site.

Check your page now

HOW IT WORKS

One address in. A plain answer out.

No account, no email and nothing to install.

  1. 01 STEP

    Enter one address

    Your home page, a landing page or a product page. No account and no email.

  2. 02 STEP

    We read it the way a crawler does

    One request for the page, one for robots.txt, and one more under each AI crawler’s name to see whether your server answers it. No JavaScript is run, and no AI service is involved.

  3. 03 STEP

    You get a plain answer

    A verdict, the words that arrived, eight checks marked pass, check or fail, and the list of crawlers.

Six crawlers, asked for by name.

For each one we read your robots.txt, then ask your server for the page under that crawler’s name.

STATIC / WORDPRESS / WEBFLOW / FRAMER

Four ways to run a site. What each one asks of you.

A fair comparison of yearly cost, upkeep time and risk, including the one place where static asks more of you.

Static filesWhat our rebuild deliversWordPressSelf-hostedWebflowHosted platformFramerHosted platform
Platform feeNoneNone. The software is freeFrom about $15 a month per siteFrom $10 a month per site
HostingAny host, from a few dollars a monthAny host with PHP and a databaseWebflow onlyFramer only
Platform and hosting a yearAbout $40 to $150About $100 to $500, with paid pluginsAbout $180 to $300About $120 to $360
Upkeep a year: updates, backups, security fixesNone. There is nothing to updateAbout $600 to $2,400 for a care plan, or one to two hours of your own time a monthNone. Webflow does itNone. Framer does it
Total to run a year$40 to $150$700 to $2,900$180 to $300$120 to $360
Updates to applyNone. There is no engine behind the pagesWordPress, theme and plugins, most weeksDone by WebflowDone by Framer
VulnerabilitiesNo login, database or plugins to attack11,334 new ones reported in 2025, 91% of them in pluginsManaged by WebflowManaged by Framer
Text in the page codeYes, alwaysUsually. Some builders and plugins add content with JavaScriptYesYes, pages are pre-rendered
Editing text yourselfBy asking an AI assistant connected to the files, or through a developer. There is no built-in editorYes, built inYes, built inYes, built in
If you leaveCopy the files to any hostMove the files and the databaseCode export on some paid plans, without CMS content or formsNo official code export

By these estimates a static site costs $40 to $150 a year to run: less than either hosted platform, and a fraction of a maintained WordPress site.

Prices as published by each provider in November 2026, in US dollars, on yearly billing, before tax. Yearly costs are our estimates for a small site, without a domain name: typical shared hosting for static files and WordPress, plus paid plugins for WordPress; twelve months of the two lowest paid site plans for Webflow and Framer; and, for WordPress upkeep, a care plan at $50 to $200 a month. Changes to the text of a static site are made with an AI assistant connected to the files, or by a developer, and are not counted. WordPress figures: Patchstack, State of WordPress Security in 2026. “Static files” means plain HTML, CSS and JavaScript, which is what our static rebuild delivers.

IF THE CHECK FINDS A PROBLEM

We fix it, at a price you can see first.

Sharpfont is an independent design and AI studio from Brașov, with 15+ years of design and product experience. Every price is published and fixed before we start.

01 / WHEN THE PAGE IS EMPTY OR THIN

Static rebuild

Your existing site rebuilt as static pages. Same design and same content, in code every crawler can read.

€450 setup, plus €30 a page

See the fixed price

02 / WHEN AI CAN READ IT, BUT IT DOES NOT PERSUADE

Pilot week

A senior design and AI team on your page for 5 working days, with a working preview by day three.

€950 one payment, no renewal

Book a pilot week

03 / WHEN THERE IS MORE TO DO EVERY MONTH

Design subscription

One team for product, brand and web. Updates every 2–3 working days. Pause or cancel before renewal.

From €3,900 a month

See the subscription

QUESTIONS

Straight answers.

Why does JavaScript matter?
Many modern sites send an almost empty page and build the text in the visitor’s browser. People never notice. The crawlers behind ChatGPT, Claude and Perplexity mostly read the page as it is sent and do not run that code, so they can receive a page with no text in it. Google’s own crawler does run JavaScript.
Does an llms.txt file help?
Not measurably today. Studies of server logs in 2026 found that almost no AI crawler requests the file, and Google says it is not needed. We do not count it in this check.
Will passing this check get me mentioned by ChatGPT?
No one can promise that. This check tells you whether AI tools are able to read your page. Whether they choose to mention you depends on what the page says and on how well known you are.
How is the score worked out?
It is out of 100. Text that arrives without JavaScript counts for 35 and crawler access for 25. The main headline and not being marked “do not index” count for 10 each, the title and description for 8, structured data for 5, real links for 4 and image descriptions for 3. A page with almost no readable text cannot score above 25, a page marked “do not index” above 40, and a site closed to every crawler above 30. The score describes this one page’s code and your server’s answers. It says nothing about how well known you are.
What if my page fails?
The result names each failed check and says why. Often the fix is small: a robots.txt rule, a missing headline or a missing description. If the page is built in the browser, the lasting fix is to send the text in the page code, which is what our static rebuild does.
Does this replace SEO?
No. It covers one thing most SEO tools do not show: what a crawler that does not run JavaScript receives. Good content and a known name still decide who gets mentioned.
My robots.txt allows everything. Why does a crawler show as “Refused”?
Some hosts and firewalls turn AI crawlers away at the server, before your site is reached, whatever robots.txt says. We ask for your page once under each crawler’s name, with our own name next to it. “Refused” means that request was turned away. A few firewalls refuse only impostors and let the real crawler through, so take it to your host as a question.
Do you keep the address I enter?
No. We request the page a few times and its robots.txt once, and show you the result. The address and the result are not stored. We do keep anonymous totals: how many checks were run, how they ended and which links on this page were used. A total cannot be traced back to you or to the site you checked. A short-lived counter limits how many checks one visitor can run, to prevent abuse.

ONE ADDRESS / A FEW SECONDS

See what AI sees.

No sign-up. Start with your home page, then the page you most want people to find.