GeoProbe
Sign inFree scan

Free · No signup · Results in ~15 seconds

Can AI search actually read your website?

ChatGPT, Claude and Perplexity only recommend products they can crawl and understand. GeoProbe checks whether yours is one of them — and tells you exactly what to fix.

We fetch four public files from your site and nothing else. Reports get a shareable link.

geoprobe.ai/scan/8xk2…Sample
57

AI readiness report

acme-saas.com

2 critical · 2 warnings · 4 passed · ~37 min of fixes

AI crawlers allowed

20 pts

Every AI crawler is blocked. ChatGPT, Claude and Perplexity cannot read your site at all.

Content in raw HTML

20 pts

Only 412 characters of text in the raw HTML — thin for AI to understand what you do.

llms.txt

10 pts

No llms.txt found. We generated one for you below.

Structured data

15 pts

JSON-LD present: Organization, SoftwareApplication.

+ 4 more checks, fix instructions and a generated llms.txt

How it works

01

Scan

We fetch your homepage, robots.txt, llms.txt and sitemap exactly like an AI crawler would — no JavaScript executed.

02

Diagnose

8 checks, each weighted by how much it affects whether ChatGPT, Claude and Perplexity can read and cite you. The weights are public.

03

Fix

Every failed check comes with the exact change to make, an effort estimate, and a ready-to-upload llms.txt.

What we check, and how much it counts

The score is a weighted sum of 8 checks totalling 100 points. No secret sauce — you can see exactly how each one moved your score.

AI crawlers allowedrobots.txt doesn't block GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot20 pts
Content in raw HTMLYour main text is server-rendered — AI crawlers don't run JavaScript20 pts
Structured dataJSON-LD for Organization / WebSite plus a page type (SoftwareApplication, Product, Article, FAQPage)15 pts
External presenceA page on Reddit, G2 or Wikipedia — where AI answers get their citations15 pts
llms.txtExists and is well-formed. Missing? We generate one10 pts
Quotable H1 and introH1 says what you are and who it's for; first paragraph answers in ≤ 60 words10 pts
Meta tagstitle, description, canonical, Open Graph5 pts
Content freshnessSomething updated in the last 3 months (sitemap lastmod or dateModified)5 pts
Total100 pts

Built for teams without a $500/month budget

Enterprise AI-visibility tools give you a polished number and no way to verify it. We do the opposite.

Transparent

Same prompt run 5 times shows “mentioned 3 of 5”, not a smoothed 60%. Every raw AI answer is one click away. Web-on and web-off are tracked separately.

Cheap, with the bill itemized

Free forever with your own API keys. Paid plans start at $29 and the pricing page shows how much of that is API cost and how much is us.

Actionable first

Monitoring alone doesn't move the needle. We start with what to fix on your site, then track whether it worked.

Tracking preview · one prompt, one model

“What are the best tools for monitoring uptime?”

hit—hithit—
3/5web search off · run weekly

Not “60%”. Five runs, three hits, five raw answers you can read. Tracking launches in the next release.

Free with your own keys. $29 with ours.

Scanning is always free. Tracking whether AI mentions you is free forever if you bring your own OpenAI / Anthropic / Google keys — you pay the providers, not us. Paid plans include API credit and show you the math.

See the full price breakdown
Free$0/mo

Bring your own API keys

1 project · 10 prompts · 2 models · weekly

Starter$29/mo

2,000 AI answers / month included

3 projects · 30 prompts · 4 models · weekly

Pro$79/mo

8,000 AI answers / month included

10 projects · 100 prompts · 4 models · daily

Start with the free scan

Takes about 15 seconds. You get a score, a fix list and an llms.txt.