See what AI can read.
Lantad is a generative engine optimization (GEO) tool. Other tools track what AI says about your brand. This one shows what AI can actually read on your pages, and what to fix first.
$ lantad scan acme-store.com
fetch as LantadBot200 OK · 41 KB
render in a real browserdone · 2.1 s
diff the two views46% missing for crawlers
robots.txt + probesGPTBot ✓ ClaudeBot ✓ PerplexityBot ✗
verdict78/100 · grade B · 3 fixes ranked
$
Renders for people. Empty for machines.
Most AI crawlers do not run JavaScript. A site can look perfect in a browser and still hand ChatGPT, Claude, or Perplexity an almost empty shell. The panels below are illustrative; your report shows your actual two views, side by side, with the numbers behind them.
Not an edge case. We measure this gap across every graded page in our live AI Visibility Index, updated as the web is scanned.
Every major AI crawler, in one scan.
Lantad checks 15 named AI crawlers and reports the robots.txt rule that applies to each, plus the live server response for every one that sends a user agent of its own. One scan covers the assistants your customers actually use.
AI assistants and answer engines
- ChatGPT
- Claude
- Perplexity
- Google Gemini
- Apple Intelligence
- Meta AI
Crawler tokens checked (15)
- GPTBot
- OAI-SearchBot
- ChatGPT-User
- ClaudeBot
- Claude-SearchBot
- Claude-User
- anthropic-ai
- PerplexityBot
- Perplexity-User
- Google-Extended
- Applebot-Extended
- Meta-ExternalAgent
- Bytespider
- CCBot
- Amazonbot
AI answers are the new front page.
More people ask ChatGPT, Claude, Perplexity, and Google's AI answers instead of scrolling search results. Those answers are built from what AI crawlers could read on your site, not from what your visitors see. If a crawler gets a blank page, your brand is missing from the answer.
You cannot be recommended if you cannot be read.
Answers, not links
AI assistants summarise instead of linking. If your content is not readable, you are not cited, and the click never happens.
JavaScript is the blind spot
Most AI crawlers do not run JavaScript. Anything that appears only after your app hydrates simply is not there for them.
A new surface, new rules
Being readable to AI is becoming as important as being indexable by Google was. Generative engine optimization needs its own checks.
Three steps, under a minute.
-
01
Fetch both ways
We request your page as an AI crawler with no JavaScript, then again in a real browser.
GET / as LantadBot/1.0 · then a full render, scrolled and settled -
02
Diff and probe
We diff the two views, read robots.txt for all 15 named AI crawlers, and test how your server answers the ones that send a user agent of their own.
token-shingle diff · robots.txt per token · one UA probe per publishing crawler -
03
Score and rank
You get a visibility score with the evidence behind every point, and a defect list, worst first.
score = 50% parity + 25% access + 15% structure + 10% schema
The weights above are the real formula, published in full on the methodology page. Nothing in the grade is a black box.
Evidence, not vibes.
AI Visibility Score
A single grade from A to F, with the sub-score and evidence behind every point.
parity · access · structure · schemaProse Parity
How much of your rendered content actually ships in the initial HTML a crawler reads. The gap between the two bars is your invisible content.
browser view vs crawler view, measured in shared textBot access matrix
Every checked crawler, from GPTBot to PerplexityBot: its robots.txt rule, plus the live server response for the ones that send their own user agent.
15 crawlers · the exact rule that decides · probe resultRanked fixes
Each defect paired with the specific change that removes it, ordered by impact, written for the person who will make the change.
severity · evidence · the exact remediationWhat AI actually reads
The extracted text itself, so you can read your page the way a crawler receives it, next to a screenshot of what a browser rendered.
extracted text · render screenshot · framework detectedShareable by design
A stable report link with a branded preview, ready for Slack, an email thread, or a client. No login needed to view it.
unguessable link · social preview card · no signup to readEvery finding links to its evidence: the exact robots line, the exact missing text, the exact server response. If we cannot measure it, the report says so instead of guessing.
Understand, prove, and keep your AI visibility, all in one place.
Most tools tell you what a chatbot said today. Lantad measures the thing underneath that decides whether AI can read and cite you at all, and shows you exactly how to fix it.
Evidence, not vibes
Every point in your score links to the thing that caused it: the exact robots.txt line, the exact text a crawler could not see, the exact server response. Nothing is a black box, and you can hand any finding to a developer as-is.
The cause, not the symptom
AI can only cite what its crawlers can read. Other tools report the symptom, whether a chatbot happens to mention you. Lantad measures the cause: how much of your page actually reaches the crawler, and whether it is allowed in at all.
Honest by design
We name the exact model behind every answer, and a page we cannot grade fairly says so instead of inventing a number. We measure what crawlers read; we never sell rankings or promise a citation. That is the whole brand.
Built for the people who own AI visibility.
One site or a hundred, in-house or on the agency side: Lantad tells you what AI crawlers can read and exactly what to fix, in plain evidence.
The fear: a buyer asks ChatGPT about you, and the answer is wrong or empty.
- See exactly what AI can read on your key pages, in plain language
- Every fix written for the person who will actually make the change
- Weekly re-scans with an email the moment anything slips
- What AI says about your brand, tested against a real model
The question you get: "can you prove this work moved anything?"
- A dated, client-facing report link per site, with no controls or billing on it, ready to forward
- Before-and-after diffs between any two checks: the proof page for the invoice
- Every client's pages grouped by site on one dashboard, re-scanned weekly
- Drift alerts into your channel and the client's, at the same time
The brief landed on your desk: "what is our GEO strategy?"
- A per-page, per-crawler readout: the baseline document GEO work starts from
- Whole-sitemap audits, every page graded worst first, with a CSV
- Prompt tracking: the questions your pages could answer, and whether models name you
- Score history that shows whether the work is working
The blind spot: the render looks fine, and the crawler view shipped empty.
- The exact missing text and the exact robots line, not a vague warning
- Scans from code over a small REST API
- An MCP server, so Claude or ChatGPT can run the check for you as a tool
- Per-framework fix hints for the stack you are actually on
Sites drift. Be the first to know, not the last.
A redeploy on Friday, a CDN rule on Tuesday, a theme update nobody mentioned: AI readability drops silently, and every AI answer about you is wrong until someone notices. Monitoring is the part of Lantad that does the noticing, every week, on every page you track.
- An AI Visibility dashboard with your average score, grade, and a trend line across every page you track.
- Weekly monitoring with drift alerts: an email the moment a page drops, and the same alert in Slack when you connect a webhook. A weekly summary is on by default and you can switch it off.
- Crawler access alerts: if a bot that could read a page last week no longer can, you hear about it even when the score barely moves.
- Whole-site audits: point us at your sitemap and see every page graded, worst first, with a CSV.
- Grouped by site for agencies, with an exportable trend you can hand to a client or a boss.
Frequently asked questions
How do I check if ChatGPT can read my website?
Paste your URL into Lantad and run a free scan. It fetches your page the way ChatGPT's crawler does, without running JavaScript, then compares that against a full browser render. You see exactly what ChatGPT, Claude, Perplexity, and other AI assistants can extract from your page, and what is missing.
What is an AI visibility scanner, and how is it different from an SEO tool?
An AI visibility scanner checks whether AI crawlers and assistants can actually read your website. Classic SEO tools check Google, which renders JavaScript. Lantad checks the AI layer: how much of your content reaches AI crawlers that do not run JavaScript, whether your server blocks them, and how well your structure and structured data survive in the raw HTML. The full method is on the methodology page.
Which AI crawlers and assistants does Lantad check?
Lantad checks 15 named AI crawlers across OpenAI (ChatGPT), Anthropic (Claude), Perplexity, Google (Gemini), Apple, Meta, ByteDance, Amazon, and Common Crawl. For each one, the report shows the robots.txt rule that applies. Crawlers that publish a request user agent are also probed with that user agent, and the report shows how your server answered. The rest are robots-only tokens, with no user agent to test.
Why can AI crawlers not read my site?
AI crawlers like GPTBot, ClaudeBot, and PerplexityBot do not execute JavaScript. If your content is rendered client-side, they receive an empty shell even though the page looks fine in a browser and in Google.
Is optimising for AI the same as Google SEO?
No. Google renders JavaScript; most AI crawlers do not. A site can rank well on Google and still be invisible to ChatGPT or Perplexity. Optimising for AI answers is often called generative engine optimization, or GEO, and it needs its own checks. That is what Lantad provides.
What is Prose Parity?
Prose Parity is how much of your rendered, human-visible content actually ships in the initial HTML that an AI crawler reads. One hundred percent means a crawler sees everything your visitors see. A low score means content is arriving via JavaScript that most AI crawlers never run.
Do I need an llms.txt file?
llms.txt is an emerging convention for pointing AI systems at your most important content. Lantad checks whether you serve one and reports its position. It is not required, but it is a low-cost signal, and the report tells you exactly where you stand.
Does Lantad spoof crawler identities?
The baseline fetch always identifies itself honestly as LantadBot. Per-bot probes send each bot's published user agent string once to detect blocking, with the results caveated, because servers that verify crawler IPs behave differently for the real crawler. Site owners can opt out.
Is the scan free, and how long does it take?
Yes, the scan is free and needs no signup. Most scans finish in under a minute. You watch the progress live and get a shareable report link at the end.
Find out what AI leaves on the table.
One URL, one scan, the evidence in under a minute.