METRIC VAULT
100% Free · No signup

Free AI Search Visibility
Checker

See whether ChatGPT, Perplexity, Gemini and Google AI Overviews can read, understand and cite your site. Get a 0-100 readiness score and the exact fixes, in seconds. It is the quickest way to check your ChatGPT visibility before changing anything. This AI search visibility checker scores one page; to track how the engines answer over time, see the AI visibility tool.

No signup · No credit card · Tests the live page, robots.txt and llms.txt
AI VISIBILITY CHECKER

See your site the way an AI engine does.

Paste one URL and get a 0-100 readiness score for whether ChatGPT, Perplexity, Gemini and Google AI Overviews can reach, read and quote your page, with the exact fixes ranked by the points they win back.

Reads the live page, robots.txt and llms.txt. No signup, no crawl setup.
ai-visibility · example
PARTLY READY
AI readiness
0
weighted, 0-100
Signals passed
0
of 11 AI checks
PASS
Crawler access
Allowed
GPTBot, ClaudeBot, PerplexityBot
20 PTS
Top fix
Add schema
no JSON-LD found yet
The problem

An AI can quietly leave you out of the answer.

When a buyer asks an assistant which tool to pick, it names a shortlist and they act on it. If the engine could not read your page, you were never in the running, and nothing tells you it happened. This check reads the signals those engines use, so you can see the gap before a rival takes the answer.

Blocked without knowing itA line in robots.txt can shut GPTBot or PerplexityBot out of your whole site, and your analytics never mention it.
Google rank tells you nothingYou can sit at position one and still be absent from every AI answer. The two do not move together.
The signals are buriedSchema, headings and llms.txt decide whether a model can quote you, and they are hidden in your markup where you never look.
How it works

One URL in. A score and a fix list out.

Paste your URL

No login and no setup. Enter your homepage and run the check.

We fetch three things

The live page, your robots.txt and your llms.txt, then inspect the eleven signals AI answer engines rely on to read and cite a site.

Get your score and fixes

A weighted 0-100 readiness score, a pass or fail on every signal, and the fixes ranked by the points each one wins back.

Glossary

What these terms mean

AI answer engine

An assistant that answers a question directly instead of listing links. It reads pages and names sources inside the answer.

ChatGPT, Perplexity, Gemini, AI Overviews

AI readiness score

One weighted number from 0-100 for how easy your page is for an engine to read, parse and quote. Higher is better.

68 = partly ready

AI crawler

The bot each engine sends to fetch your page. If robots.txt blocks it, that engine never sees your content at all.

GPTBot, ClaudeBot, PerplexityBot

Structured data

Hidden schema.org labels in your code that tell an engine what your page is about, so it does not have to guess.

JSON-LD, worth 20 points here

FAQ structure

Content shaped as questions and answers. Answer engines lift it almost whole, so it is one of the easiest ways to get quoted.

FAQ or Question schema

llms.txt

A small text file at your root that points AI crawlers to the pages that matter most, the way a sitemap guides search engines.

served at /llms.txt
One clear number

A readiness score you can act on.

Every signal is weighted and rolled into a single 0-100 score, so the first thing you see is whether an AI engine can work with this page today, not a wall of raw checks.

  • Weighted so heavy signals move it the most
  • Three plain bands so the number means something
  • The exact score the free checker returns

Rule of thumb: 80 and up is strong and worth protecting, 55 to 79 is partly ready with points leaking, and under 55 means an engine is likely skipping the page.

AI readiness score
0 OUT OF 100
80-100 strong55 to 79 partly ready0 to 54 at risk
Signal by signal

A pass or fail on all eleven checks.

The score is only the headline. Below it, each signal an answer engine relies on gets a clear pass or fail, so you see exactly what is working and what is holding you back.

  • Crawler access, structured data and clear titles
  • Headings, FAQ shape, canonical and Open Graph
  • llms.txt and enough real content to quote

Why it works: an engine can only name a page it can crawl, parse and trust. Each green chip is one of those hurdles already cleared.

What AI engines see on northpeakcrm.com
Crawlers allowedStructured dataTitle tagMeta descriptionSingle H1Section headingsQuestion / FAQllms.txtCanonical URLOpen GraphContent depth

Eight green, three grey. 8 of 11 passed, which is why the score lands at 68 and not higher.

Ranked by impact

A fix list ordered by points won back.

Every failed signal becomes a fix, sorted by the points it recovers, so you always know the single change that moves the score most. Work the list from the top down.

  • Only the checks you actually failed, nothing generic
  • The point value on each, so effort follows reward
  • A plain instruction, not a raw error code

Pro tip: clear the heaviest fix first. Adding schema.org alone lifts this example from 68 to 88, straight from partly ready into strong.

0
points to recover, ranked
Add structured data
+20 pts
Add a Q&A or FAQ section
+8 pts
Publish an llms.txt file
+4 pts
See it in action

How to read a result to a plan.

A score only helps if you know what to do with it. Here is the exact read on our example, northpeakcrm.com, the way an experienced SEO would think it through.

68/100Readiness
Partly ready, not absent. The foundations are there, but three signals are leaking points to a competitor an engine understands better.
YesCrawlers allowed
The door is open. GPTBot, ClaudeBot and PerplexityBot can all reach the page, so every other signal actually counts. This is the one you never want to fail.
0Schema, 20 pts
The biggest gap. With no JSON-LD, the engine has to guess what this company is. Add schema.org and you win back 20 points, more than any other single move.
NoneFAQ, 8 pts
Nothing to lift. There is no question and answer content, the shape answer engines quote most. A short FAQ section recovers 8 points and gives them a ready made answer.
100After 3 fixes
Add structured data, an FAQ section and an llms.txt file and this page climbs from 68 to a perfect 100, from partly ready to fully quotable. That whole plan came from one scan.
Where it shines

Built for teams that want the answer.

Founders

Know if you are ready to be named

See whether your site gives the assistants what they need: allowed crawlers, clear labels and quotable content, before you spend on marketing.

SEO and marketing leads

Audit a surface Google cannot see

Get a fast read on AI readiness and the exact fixes for a channel classic rank tracking misses entirely, then report on it.

PR and content teams

Turn every failed signal into a brief

Each grey chip becomes a clear instruction for the next page, so the engines repeat your story instead of a guess.

Agencies

Score a prospect in seconds

Run the check before a pitch and walk in with a credible, ranked fix list instead of a vague promise about AI search.

Good to know

Questions, answered.

No, and it is worth being clear about that. This reads your live page, robots.txt and llms.txt and scores the on-page and technical signals AI engines rely on to read, parse and cite a site. Tracking whether ChatGPT, Gemini, Perplexity or Claude actually name you over time is the paid Metric Vault dashboard.

It is a weighted 0-100. Crawler access carries the most weight at 22 points, then structured data at 20, title and meta description at 12 each, headings and FAQ shape at 8 each, with llms.txt, canonical and Open Graph filling the rest. Fail a heavy signal and the score drops the furthest.

Rank tracking tells you where you sit in the ten blue links. It says nothing about whether an assistant names you in its answer, and the two do not move together. You can rank first and still be absent from every AI answer, so AI visibility needs its own check.

Work the ranked fixes from the top. Most low scores share the same causes: an AI crawler blocked in robots.txt, missing structured data, or a page too thin to quote. Allow the crawlers first, add schema.org, then rewrite key pages to say plainly what you are and who you serve.

Yes. Every signal comes with a fix attached, and the list is ordered by impact. Paste your URL and follow the list from the top.

The signals it checks, crawler access, structured data, clear content and headings, are the same ones all the major answer engines rely on to read and cite a site. Get those right and you are readable to every one of them at once.

Ready when you are

See your site the way an AI engine does.

It takes one URL. You get a 0-100 readiness score, a pass or fail on all eleven signals, and the fixes ranked by the points they win back.

See if AI actually recommends you

The checker tells you if AI can read your site. Metric Vault tracks whether ChatGPT, Gemini, Perplexity and Claude actually mention you, how you stack up against competitors, and how it trends over time, plus the rest of the platform's 60+ SEO and AI tools.