T&B Technology
Start a project

FREE TOOL

AI visibility check: can AI assistants read your site?

Type in your address. We read the page and the files meant for machines the way any crawler reads them, and in about ten seconds you see whether ChatGPT, Claude and Perplexity can read you and tell who you are.

The home page or any other page.

We read the page and three public files meant for machines, the way any crawler does. We identify ourselves as TnB-SiteCheck and never pretend to be another crawler. The address is not kept; the result stays for 10 minutes.

Result

Assistant crawlers

robots.txt for OAI-SearchBot, Claude-SearchBot, PerplexityBot, Bingbot, Googlebot and 8 more

—
Training crawlers

robots.txt for GPTBot, ClaudeBot, Google-Extended and 7 more

—
Text without JavaScript

the words in the first HTML, before any script runs

—
Who you are

Organization or LocalBusiness: name, address, phone, profiles

—
Snippet blocks

noindex, nosnippet, max-snippet in meta robots and X-Robots-Tag, data-nosnippet

—
Freshness

the newest date (lastmod) in the sitemap

—
llms.txt

/llms.txt

—
Bing

Bing Webmaster Tools verification (msvalidate.01)

—

What this means

WANT A HAND

Send us the result

This is where our SEO and AI search work starts: an audit, then the technical work, the content and the structured data.

Replace your text with the result?

ENQUIRY RECEIVED

We have it. We reply within two working days.

A copy is on its way to your inbox. If it does not arrive, check the spam folder — and write to us directly. contact@tnb-tech.com

Reference number:

Which layer hurts?

Pick the service you need — the small mark beside it is our name for that layer in the story.

The text below comes from your result. Read it, change it or delete any of it. Nothing is sent until you press Send.

A brief, a screenshot or a sketch. Up to 8 MB.

Picked: Marketing & search

We use what you write only to answer you. See the privacy policy.

How to fix it

robots.txt: search allowed, training blocked

These lines let in the crawlers that search and cite, and block the ones that collect text to train models. Put them in the robots.txt at the root of your site.

# Search and answers: allowed. Copy the Disallow lines from your
# User-agent: * group here as well: a crawler named here skips that group.
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
User-agent: Claude-SearchBot
User-agent: Claude-User
User-agent: PerplexityBot
User-agent: Perplexity-User
User-agent: Google-NotebookLM
User-agent: DuckAssistBot
User-agent: Applebot
User-agent: meta-externalfetcher
User-agent: MistralAI-User
Allow: /

# Training of language models: blocked. This does not stop citations.
User-agent: GPTBot
User-agent: ClaudeBot
User-agent: Google-Extended
User-agent: Applebot-Extended
User-agent: meta-externalagent
User-agent: Amazonbot
User-agent: CCBot
User-agent: cohere-ai
User-agent: Diffbot
User-agent: Omgili
Disallow: /

Business facts (JSON-LD)

Put the block in the head of your home page. The website address is filled in; add the name and the facts, and delete any line you leave empty.


            

                
                
            

        


        
        

    

        


                        

            
What we check

        
We read the page and three public files meant for machines (robots.txt, the sitemap and llms.txt), the way any crawler reads them. We do not ask an assistant: one answer is one random sample. We check what decides whether it can cite you at all.

        
We only look at what the site shows every visitor and at the domain's public records. We do not try passwords, forms or hidden addresses.


                    
This is where our work on SEO and AI search starts: an audit, then the technical work, the content and the structured data.

        
                    

                In the glossary:
                                    llms.txt ·                                     Generative engine optimisation (GEO)                            



        
        
                    

                

                    

            
QUESTIONS

                
What people ask about AI visibility

            



                    

                                                    

    
                    
What is AI visibility?

            
    
Whether ChatGPT, Claude, Perplexity and the other assistants can read your site, tell which business it is and name it in an answer. The first part depends on robots.txt and on the text in the HTML. The second depends on clear facts about the business, on your site and on other sites.



                                                    

    
                    
How does ChatGPT decide which businesses to recommend?

            
    
The exact rule is not public, but three things are plain. An assistant only recommends sites it can read. It prefers pages with clear facts: name, address, phone, services. And it trusts a business that other sites mention: in an Ahrefs study of 75,000 brands, mentions were the strongest link to showing up in AI answers.



                                                    

    
                    
Should I block GPTBot?

            
    
That is your choice. GPTBot collects text to train future OpenAI models. Blocking it does not take your site out of ChatGPT: search goes through OAI-SearchBot, and opening a page because a person asked goes through ChatGPT-User. If you want to be cited, do not block those two.



                                                    

    
                    
What is llms.txt, and do I need one?

            
    
A text file at the root of a site that tells language models which pages matter and what is in them. It is optional and not a standard: Google says it does not use it, and some assistants and agents read it. It costs little, but robots.txt and the text of your pages count for more. This page runs an llms.txt check too: you see whether the file is there, and whether it is text rather than an HTML page.



                                                    

    
                    
Why don’t you just ask ChatGPT?

            
    
Because one answer is one random sample. Ask the same question a minute later, or as someone else, and you get a different list. We check what decides whether an assistant can cite you at all, and what you can change.



                                                    

    
                    
Does structured data help?

            
    
It helps machines tell who you are without guessing: your name, address, phone and profiles sit in one block they read directly. Google says it is not required for AI search; Microsoft says it helps its models understand content. Without it, an assistant has only your text and other people’s sites to go on.



                                            

                

            

        
                    

        

            

                
                                    All tools
                            

            

                                    

    

        in your browser
        
ROAS calculator

                    
Is your advertising making money? ROAS, break-even and the most you can pay for an order, from your own margin.

            

    ROAS calculator


                                    

    

        in your browser
        
Ad budget calculator

                    
A monthly and a daily budget worked out from your goal, your cost per click and your conversion rate.

            

    Ad budget calculator


                                    

    

        on our server
        
Indexing check

                    
noindex, robots.txt, canonical, the sitemap and redirects: the usual reasons Google leaves a page out.

            

    Indexing check