Home › AI visibility audit
Can ChatGPT, Claude and Perplexity read your website?
Quick answer
The AI visibility audit tells you, in plain English, whether the crawlers behind ChatGPT, Claude, Perplexity, Google and Bing can reach your pages, read them without running JavaScript, and understand what you sell. You get a PDF report within 48 hours and a fix pack of ready-to-paste files. $79 for the report, $149 with the fixes applied. It is a diagnosis: nobody can guarantee that an AI assistant will cite or recommend you.
Order an audit
Report + fix pack. PDF report and ready-to-paste files you or your developer apply.
Pay $79 and we startReport + fixes applied. Same report, and we make the changes for you.
Pay $149 and we startPay and we start. Secure checkout by Stripe asks for your website address. Your PDF report arrives within 48 hours of payment, covering your home page and up to 5 more pages; reply to the confirmation email to tell us which pages matter most. Questions? hello@astorventuresllc.com
What the audit checks
Each check runs against your home page and up to 5 more pages you pick (for example your main product, service or pricing pages). The crawlers covered are GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Google-Extended, Googlebot, Bingbot, Applebot-Extended, CCBot, Meta-ExternalAgent, Amazonbot and DuckAssistBot.
| Area | Points | What is checked |
|---|---|---|
| AI crawler access in robots.txt | 25 | Whether your robots.txt lets each of 16 crawlers read the page, using the same group and longest-match rules crawlers follow (RFC 9309). |
| Live fetch as each crawler | 15 | Status code, size and challenge pages when the page is requested with each crawler’s user agent, so firewall and CDN blocks show up. |
| Content visible without JavaScript | 20 | Words of text in the raw HTML, whether the main content is there, and signs of an empty JavaScript app shell. |
| Structured data | 15 | JSON-LD types (Organization, Product, Article, FAQPage and so on), invalid blocks, dates, author and sameAs. |
| Title, description, canonical, H1, indexable | 15 | The basics every crawler reads first, plus noindex in meta robots or the X-Robots-Tag header. |
| Sitemap | 5 | Whether a sitemap exists and is listed in robots.txt. |
| Citation readiness | 5 | Question-style headings with answers under them, dates, and enough text to quote. |
| llms.txt | 0 (noted only) | Reported as present or missing. Not scored, because no major AI company documents that its crawlers use it. See Does llms.txt work? |
On top of the automated checks, we read the pages the way an assistant would: is there a short, direct answer to the questions your customers ask, are prices, service areas and contact details in plain text, and does the schema match what the page says?
What you get
- A PDF report within 48 hours: a 0–100 score per page, findings ordered by priority in plain English, and a crawler-by-crawler table (robots.txt rule, status code, bytes). See a sample report (PDF).
- A fix pack of ready-to-paste files: a robots.txt block for the AI crawlers you choose to allow or block, JSON-LD for your business and key pages (including FAQPage only where the page really has questions and answers), suggested titles and meta descriptions, and a short checklist for your developer or CDN settings when a firewall is the problem.
- With the $149 option: we apply the fixes for you, with temporary access to your site builder or hosting, or by sending exact changes to your developer.
- llms.txt: if you want one we can draft it, but we list it as optional. It is not the fix.
Sample findings from real audits
| Site (home page) | Score | What the checker found |
|---|---|---|
| astorventuresllc.com | 98/100 | All 16 crawlers allowed; 3,664 words in the raw HTML. Small gaps: a 253-character meta description and no sameAs profile links in the schema. |
| hohocam.com | 94/100 | All 16 crawlers allowed; 2,165 words in the raw HTML; Product, Offer, FAQPage and VideoObject schema. Gaps: no <main> element and no published or updated date in the schema. |
| kemuva.com | 87/100 | All 16 crawlers allowed; 2,809 words in the raw HTML. Gaps: only Organization and WebSite schema on the home page, one question-style heading, no <main> element, no dates in the schema, and a 193-character meta description. |
Run the free checker yourself
The checks above are automated in a single Python file that you can read before you run it. It uses only the Python standard library: no installs, no API keys, nothing sent to us.
Download ai_audit.py Optional: HTML/PDF report maker
- Install Python 3.8 or newer (Mac and Linux usually have it; on Windows get it from python.org).
- Open a terminal in the folder where you saved the file and run:
python3 ai_audit.py https://yourbusiness.com --out reports - Open
reports/yourbusiness.com.mdfor the readable report, or.jsonfor the raw data. To get a PDF, runpython3 make_report_html.py reports/yourbusiness.com.json report.htmland print the HTML page to PDF from any browser.
How it works, step by step
- robots.txt: downloads
/robots.txt, finds the group that names each crawler (or the*group if none does) and applies the longest matching Allow or Disallow rule to your page’s path. - Live fetch: requests the page once with a normal browser user agent and once with each crawler’s published user agent. A 403, 429, 503, a challenge page or a much smaller response counts as a problem.
- No-JavaScript view: strips scripts and styles from the raw HTML and counts the words left, inside and outside
<main>or<article>, and looks for app-shell markers like an empty<div id="root">. - Page signals: reads JSON-LD types, title, meta description, canonical, H1 and H2 headings,
lang, meta robots and X-Robots-Tag. - Site files: checks the sitemap (from robots.txt, or
/sitemap.xml) and whether/llms.txtexists. - Score: adds up the points in the table above and writes plain-English findings, most important first.
Limits you should know
- The requests come from your own IP address with a copied user agent. Real crawlers come from their own published IP ranges, and a CDN or firewall can treat them differently in either direction. A 200 is a good sign, not proof.
- Google-Extended and Applebot-Extended are robots.txt tokens only. They never fetch pages, so only their robots.txt rule is checked.
- It checks one URL per run. Run it on each page that matters.
- A high score means crawlers can technically reach and read the page. It does not mean any assistant will cite it.
What this audit is not
- Not a guarantee of citations, rankings, traffic or sales. Those depend on your content, your competitors and each AI company’s systems, which change often.
- Not a monthly tracking subscription. It is a one-off check you can repeat with the free script.
- Not a content-writing or link-building service.
Ready? Pick a plan above. Pay and we start; the 48-hour clock starts at payment. Questions first? Email [email protected].
Frequently asked questions
Will this get my business cited by ChatGPT, Claude or Perplexity?
No one can promise that. The audit makes sure nothing technical stops these assistants from reading your pages, and points out content gaps. Whether an assistant cites you also depends on the question, your competitors and how each AI system chooses sources.
Do I need an llms.txt file?
It is optional. llms.txt is a proposed file, and as of October 24, 2026 none of OpenAI, Anthropic, Perplexity or Google says in its crawler documentation that its crawlers use it. Google’s guidance says you do not need new machine-readable or AI text files to appear in its AI features. We report it, but we do not score it or sell it as the fix.
What is the difference between the free checker and the paid audit?
The free checker automates the technical checks for one URL at a time. The paid audit runs them on up to 6 pages, then a person reviews the results, reads your pages for answer and citation gaps, and writes the fix pack: robots.txt, JSON-LD, titles and descriptions, and developer or CDN steps.
What do you need from me?
Only your website address, which checkout asks for, and, if you like, the pages that matter most (reply to the confirmation email). For the $149 option we also need temporary access to your site builder or hosting, or your developer’s contact.
How do I pay?
Pick a plan and pay by card through secure Stripe checkout, which asks for your website address. Pay and we start: your PDF report arrives within 48 hours of payment.
Should I block GPTBot or other training crawlers?
That is a business choice. Blocking GPTBot tells OpenAI not to use your content for model training, and OpenAI says it is a separate setting from OAI-SearchBot, which controls ChatGPT search. The report shows what you allow today and gives you a robots.txt block for whichever choice you make.
Does it work for Shopify, Wix, Squarespace or WordPress sites?
Yes. The checks only need a public URL. Where the fix is a setting in your platform or CDN rather than a file, the fix pack says where to change it.