Free tool · No sign-up
Free AEO checker
An AEO checker tests whether a page is ready to be quoted by answer engines such as ChatGPT, Perplexity and Google AI Overviews. Paste any URL below: Placton’s free AEO checking tool runs 17 checks, from AI crawler access to a direct answer under the H1, and tells you what to fix first.
7 days free. No card required.
Placton tracks
ChatGPT
Claude
Google AI
Perplexity
Grok
Run a free AEO check on any page
Paste the address of an article, a product page or your homepage. We read it the way an AI crawler does and score it.
Free and without sign-up. Paste any page: an article, a product page or your homepage.
What our AEO checking tool tests: 17 checks
Grouped as in your result. A pass scores 1, a warning half and a fail 0, and the score is the average out of 100. If the page fails a check that stops it being cited at all, marked as a blocker, the score is capped at 40.
Access5 checks · Can answer engines reach and read this page?
Blocker
Answer engine crawlers allowed
Test: robots.txt lets OAI-SearchBot (ChatGPT search), Googlebot (Google AI Overviews), Claude-SearchBot and PerplexityBot read this exact path. One blocked is a warning; Googlebot, OAI-SearchBot or any two blocked is a fail.
Why it matters: Training crawlers such as GPTBot are reported but do not count: blocking them keeps your pages out of model training, not out of answers, as OpenAI’s crawler documentation (opens in a new tab) explains. The AI crawler checker lists 16 crawlers with the exact line.
Blocker
Indexable and quotable
Test: No noindex, nosnippet or max-snippet:0 in the robots meta tag or the X-Robots-Tag header. A snippet limit under 160 characters is a warning.
Why it matters: Google applies these rules to its AI features as well, so a page that may not be quoted cannot become the quoted answer.
Blocker
Text served without JavaScript
Test: 300 words or more of main text in the HTML the server sends is a pass, 80 to 299 a warning and fewer a fail.
Why it matters: Most AI crawlers do not run JavaScript, so text that scripts add later is invisible to them.
Canonical URL
Test: A canonical tag that points to this page. A missing canonical, or one that points to another address, is a warning.
Why it matters: Search and answer engines cite the canonical version, so a canonical that points elsewhere sends the credit there.
llms.txt on the site
Test: A text file at /llms.txt on the same host. None, or a web page in its place, is a warning.
Why it matters: A proposal rather than a rule, and cheap to add: the llms.txt generator drafts one from your sitemap.
Answer first4 checks · Does the page answer the question up front?
Direct answer at the top
Test: The first paragraph after the H1, before the first section, has 25 to 90 words and does not open with a preamble such as “Welcome to”. Bylines, dates, tables of contents and a download box in between are skipped. A summary list of 15 to 120 words passes too. No opening paragraph at all is a fail.
Why it matters: Featured snippets and AI answers tend to lift a self-contained passage of about 40 to 60 words from near the top.
Question-style headings
Test: At least two H2 or H3 subheadings phrased as a question or starting with how, what or why, and at least half of them answered in the first paragraph below, in 90 words or fewer.
Why it matters: People ask answer engines questions, and a heading that matches one marks where its answer starts.
Heading structure
Test: Exactly one H1, at least two H2 sections and no skipped levels, such as an H2 followed by an H4. No H1 is a fail.
Why it matters: Clear sections let a machine lift the part that answers a question without the rest of the page.
Lists or tables
Test: At least one list of two items or more, or a table, in the main content.
Why it matters: Steps, options and comparisons are often quoted straight from lists and tables.
Content3 checks · Is there enough to quote, and is it easy to lift?
Enough depth
Test: 500 words or more of main text is a pass and 250 to 499 a warning. Fewer is a fail, unless the page is almost empty, which the JavaScript check already counts.
Why it matters: A page that covers the follow-up questions has an answer for more of the ways people ask.
Easy to read
Test: Sentences average 20 words or fewer, at most 15% run past 30 words and paragraphs average 90 words or fewer. Up to 26 words a sentence is a warning. English pages also get a Flesch reading ease score, for reference.
Why it matters: Short, plain passages stand on their own when they are quoted out of context.
Links to sources
Test: The main content links to at least two outside sites, not counting social networks and share buttons. One is a warning.
Why it matters: Linking each fact to its source shows where your figures come from, the same habit that makes a page worth citing.
Trust2 checks · Who says it, and is it current?
Named author
Test: An author in the structured data or the author meta tag, a rel="author" link, a link to an author page or a “By” line near the top.
Why it matters: A named person who stands behind the page is easier to trust than an anonymous one.
Published and updated dates
Test: A publication or update date in the structured data, the article meta tags or a <time> tag, from the last two years.
Why it matters: For questions whose answer changes, a date is how a reader knows the answer still holds.
Structured data3 checks · Can a machine tell what the page is?
Structured data for the page
Test: Valid JSON-LD or microdata with a type that says what the page is, such as Article, Product, HowTo or LocalBusiness, not only WebPage or BreadcrumbList. Invalid JSON is a warning.
Why it matters: Google says AI Overviews need no special markup, but schema.org states facts such as author and dates without leaving them to interpretation.
FAQ or Q&A markup
Test: FAQPage markup with at least one question, or QAPage markup.
Why it matters: Google shows FAQ results for few sites since 2023, but the markup still pairs each question with its answer in a form machines read directly.
Title and description
Test: A title of 10 to 65 characters and a meta description of 50 to 170. No title is a fail.
Why it matters: The title is the name an answer engine shows next to the link it cites.
How to pass an AEO check: one page, before and after
An invented roofing company with made-up prices. Same service, same site: the second version puts the answer first, asks the questions customers ask and says who stands behind it.
Before
Score 68/100
H1
Roof replacement
First paragraph
Welcome to Example Roofing! Since 1998, our family-owned team has proudly served homeowners across the region with quality, integrity and a commitment to excellence on every job, big or small.
H2 headings
- Our process
- Why choose us
- Get in touch
No author, no date, no list, no outside source and no FAQ. Organization markup only.
- !Direct answer at the top: opens with a preamble
- !Question-style headings: none of 3
- !Named author and dates: not found
- !Structured data: only general types
After
Score 94/100
H1
Roof replacement cost in Austin
First paragraph
Replacing the asphalt shingle roof of a typical 2,000 square foot home costs $12,000 to $18,000 with us and takes two to three days. Metal costs about twice as much and lasts about twice as long. Size, pitch and the layers to remove set the final price.
H2 headings
- How much does a new roof cost?
- How long does a roof replacement take?
- Do I need a permit to replace my roof?
A named author, publication and update dates, a price list, two official sources, and Article, LocalBusiness and FAQPage markup. No llms.txt yet.
- Direct answer at the top: 47 words
- Question-style headings: 3 of 3, each answered below
- Named author and dates: found
- Structured data: Article, LocalBusiness and FAQPage
Nothing in the second version is a trick. It answers the question a customer types, in the words they use, near the top, and backs it with a name and a date. That is also what makes a page easy for a person to scan.
AEO checking tools: page check, site audit or answer tracking?
Three kinds of tool hide behind the same words. They answer different questions, and most sites need all three at some point.
Page check
Can answer engines read and quote this page?
- Looks at
- One URL: its robots rules, its HTML without JavaScript, its opening answer, headings, author, dates and markup.
- When
- Before you publish, after a rewrite and on every page you want cited.
- Cost
- Free on this page.
Site audit
Is the whole site open to AI crawlers?
- Looks at
- The homepage, robots.txt, the sitemap and llms.txt, in 18 site-wide checks.
- When
- Once, and again after changes to the site or its hosting.
- Cost
- Free, no account.
Answer tracking
Do the AI answers actually name us?
- Looks at
- Real answers from ChatGPT, Google AI Overviews, Claude, Perplexity and Grok to the questions your customers ask, with the pages they cite and the rivals they name.
- When
- Every day and every week, because answers change.
- Cost
- Placton plans from $59 a month, 7 days free.
What the best AEO checker should show you
It reads what a crawler gets. The HTML the server sends, before any script runs, and not the page your browser paints.
It checks the page, not only the homepage. robots.txt rules can close one folder and leave the rest open, so the test has to use the exact path.
It separates answer crawlers from training crawlers. Blocking GPTBot is a choice about training. Blocking OAI-SearchBot keeps you out of ChatGPT search.
It shows the text it found. The title, the opening answer and the question headings, so you can judge what would be quoted.
It explains every threshold. A score you cannot trace back to a rule is hard to act on.
It is clear about its limits. A ready page is not a cited page. Only the answers themselves show whether you are named.
We make Placton, so weigh our view accordingly. Google’s own guidance on AI features and your website (opens in a new tab) is a good second opinion.
A ready page is not a cited page
The check tells you whether anything on the page stands in the way. It cannot tell you whether an answer engine names your brand when a customer asks. Placton sends those questions to ChatGPT, Google AI Overviews, Claude, Perplexity and Grok, and shows when your brand appears, which pages are cited and who is named instead.
What the trial includes
The check covers one page. The trial covers the answers: what five AI platforms say when your customers ask.
7 days free, no card. Sign up with your email or Google. When the trial ends, tracking stops and nothing is charged.
Up to 10 prompts. We suggest prompts and competitors from your website, and you edit the list before the first run.
All five AI platforms on day one. ChatGPT, Google AI Overviews, Claude, Perplexity and Grok answer your prompts as soon as your email is confirmed.
ChatGPT and Google AI again every day while you keep visiting your dashboard. Claude, Perplexity and Grok are asked once, at sign-up.
One “Run now” and the 18-point site audit. Anyone can also run it free in our GEO audit tool, and check single pages here.
On paid plans only: team invites and the weekly email summary with alerts.
* On paid plans. ChatGPT is read on chatgpt.com in every country Placton tracks except Venezuela and Puerto Rico, where it is asked weekly through its API.
Placton Starter
Track 20 prompts on 5 AI platforms
$59/ month (USD)
Or $602 billed yearly · save $106
Start your 7-day free trial7 days free. No card required.
- Up to 20 prompts, about 1,440 answers a month
- ChatGPT, Google AI Overviews, Claude, Perplexity, Grok
- ChatGPT and Google AI every day*, the other three every week
- Competitor ranking (up to 25 rivals) and every cited URL
- Weekly insights, plus an email summary with alerts
- 18-point AI-readiness site audit
- Up to 2 team members
Need more prompts? Growth is $129 a month for 75 and Scale $299 for 250. Compare all plans
AEO checker FAQ
About Placton itself? Read the full FAQs or see every plan and price.
What does an AEO check look at?
Whether one page is ready to be quoted by answer engines: if their crawlers may read it, if its text is in the HTML, if it answers the question at the top, if its headings match the questions people ask and if it says who wrote it and when. This tool runs 17 checks and scores them out of 100.
Is this AEO checker free?
Yes, with no account, on any public page, yours or a competitor’s. Each connection gets 10 checks an hour and 30 a day, and neither the address nor the page is stored.
What is a good AEO score?
Above 80 means nothing blocks the page and only small gaps are left. Between 50 and 80, work through the Fix first list. Below 50, or a score capped at 40, something stops answer engines from reading or quoting the page: fix that before anything else.
How is it different from the free GEO audit?
The GEO audit looks at the site from its homepage, with 18 site-wide checks on robots.txt, the sitemap and llms.txt. This check looks at one page you choose: its own robots rules, its opening answer, its questions, its author and its dates. Audit the site once, then check each page you want cited.
Will a high score get my page cited by ChatGPT?
No tool can promise that. A high score means nothing on the page stands in the way. Whether an answer engine picks it depends on the question, on the other pages it finds and on what other sites say about you.
The only way to know is to ask the AI and read the answer, which is what Placton’s plans do every day and every week.
Why does it say my page has almost no text?
We read the HTML your server sends without running JavaScript, as most AI crawlers do. If your site builds its pages in the browser, that HTML is close to empty. Server-side rendering or static pages fix it.
Should I block GPTBot to protect my content?
That is your decision, and it does not change this score. GPTBot collects data that may train OpenAI’s models, while OAI-SearchBot finds pages for ChatGPT search. Blocking the first keeps your pages out of training; blocking the second keeps you out of ChatGPT’s search answers.
Does it check Google AI Overviews?
It checks what Google needs before it can use a page there: Googlebot may crawl it, it can be indexed and no nosnippet rule stops Google from quoting it. Google says AI Overviews need nothing beyond that. Whether your page shows up is something only the live results can tell.
Your page is ready for answer engines. Are you in the answers?
Enter your website and Placton drafts the questions your customers would ask and the competitors to watch. You edit both, and the first answers arrive the same day.
7 days free. No card required.