Flisko
Free tool

AI visibility checker. Can ChatGPT, Claude and Perplexity actually read your site?

An AI visibility checker tests whether AI search engines can read, understand and cite a website. This one reviews twenty-one technical signals, scores them from 0 to 5, adds an independent agent-readiness score from 0 to 100, and emails you the report within minutes. Free, no sign-up, any site — including your competitors'.

✦ Free AI visibility test
  • Free
  • No obligation
  • Report by email within minutes
Price
free, no limits
Signals
21 technical checks in 5 groups
Scores
level 0–5 and agent readiness 0–100
Report
by email in a few minutes
Engines
ChatGPT, Claude, Perplexity, Gemini

What the AI visibility checker tests

An AI search engine does not return ten links for you to choose between — it returns one answer and names one or two companies in it. To be one of them, the model has to be able to read your site first. The checker tests exactly that, in five groups. The first three decide whether you can be cited at all; the last two are advanced standards for AI agents.

The five groups of signals the checker tests, and what a failure means
GroupWhat is checkedWhat a failure means
1. Discoverabilityrobots.txt, sitemap, link headers, DNS recordsCrawlers cannot find your pages, or find only some of them.
2. Content accessibilityWhether the text is readable without running a browserThe page renders only in JavaScript; AI crawlers see an empty shell.
3. Bot access controlWhich AI crawlers are allowed, and for whatGPTBot, ClaudeBot or PerplexityBot are blocked or left unnamed.
4. Agent service discoveryStandards an agent uses to connect to your servicesAdvanced; a next step for most businesses, not a shortcoming.
5. Commerce protocolsBuying and paying through agentsOnly relevant for online shops.

Both scores are computed by independent, publicly available scanners, not by us. We run them, combine the two reports and send them in one email — so the result is verifiable rather than something we made up.

How to read your score

  • Level 0–2: at least one core group fails. AI engines either cannot find your content or cannot read it. This is where most business sites built on page builders land, and it is fixable in days.
  • Level 3: the three core groups pass. The site is discoverable, its text is readable without a browser, and AI crawlers are explicitly allowed. This is the target for most companies.
  • Level 4–5: agent standards on top — service discovery and commerce protocols. Relevant for shops and software products.
  • Agent readiness 0–100: an independent second opinion on how well an AI agent could act on your site. Most well-built business sites land between 40 and 70; the report says which items move it.

What to do with a low score

The report lists the core checks your site does not pass. Most of them come down to four things: server-rendered content, an AI-crawler block in robots.txt, JSON-LD structured data about the company, and an llms.txt file. You can do them yourself — this checklist and the llms.txt template are enough for a developer — or we deliver them as a one-off AI search setup for 199 €, a final price, with a re-test and a before-and-after report.

A good score is the precondition, not the result. Whether ChatGPT actually mentions you depends on what it reads once it can read you — that is the work described on the generative engine optimization page, together with two cases with numbers and the prompt in which ChatGPT named us first.

Frequently asked questions about the AI visibility checker

  • What does the AI visibility checker test?

    Twenty-one technical signals in five groups — discoverability (robots.txt, sitemap, link headers, DNS), content accessibility (can a model read the text without running a browser), bot access control (which AI crawlers are allowed and for what), agent service discovery and commerce protocols — scored as a level from 0 to 5. The emailed report adds an independent AI agent readiness score from 0 to 100.

  • Is the AI visibility checker really free?

    Yes. You enter a web address and your email; the report arrives within a few minutes. No sign-up, no card, no subscription, and you can check any number of sites, including competitors. The address and email are used only to deliver the report and reply to it.

  • Does the checker show whether ChatGPT mentions my brand?

    No — it checks whether ChatGPT, Claude, Perplexity and Gemini can read and understand your site, which is the precondition for any mention. Brand-mention tracking, with the same prompts put to the models each month and the answers quoted verbatim, is part of our monthly monitoring service.

  • What is a good score?

    Level 3 or higher on the 0–5 scale means the three core groups pass: the site is discoverable, its content is readable without a browser, and AI crawlers are explicitly allowed. Levels 4 and 5 add agent standards that most businesses do not need yet. On the 0–100 agent-readiness score, most well-built business sites land between 40 and 70.

  • Who computes the scores?

    Two independent, publicly available scanners, not us — so the result is verifiable rather than something we made up. We run them, combine the two reports and send them to you in one email.

  • What do I do with a low score?

    The report lists the core checks your site fails. Most are fixed by four things: server-rendered content, an AI-crawler block in robots.txt, JSON-LD structured data about the company and an llms.txt file. We deliver exactly that as a one-off AI search setup for 199 €, a final price, with a re-test and a before-and-after report.

✦ Next step

Run the check, then decide.

The report is free and yours to keep. If it shows something needs fixing, the setup costs 199 € — or we talk it through on a 30-minute call first.