Free · No sign-up · Every score explained
Can ChatGPT, Google and Perplexity actually read your site?
SchemaCraft is a free AI search visibility audit. It shows whether ChatGPT, Google, Perplexity and Claude can crawl, read and cite a page, using about 50 SEO, GEO and AEO checks. Every problem includes the evidence and a copy-paste fix. The full report is free, limited to 20 audits per 10 minutes, and most audits finish within 20 seconds.
How it works
1. Real crawl, no guessing
Our server requests your page, robots.txt, sitemap and llms.txt, then repeats the request as Googlebot, OAI-SearchBot, PerplexityBot and ClaudeBot.
2. Score built only from evidence
Each check is pass, warning or fail with the raw evidence. Anything we could not verify is marked “Not assessed” and does not count.
3. Copy-paste fixes
Every issue comes with the exact robots.txt line, meta tag or JSON-LD to add — prefilled with your site’s details where we can.
Pricing
SchemaCraft pricing is zero. There is no paid plan and no sign-up. The full report is free, with a limit of 20 audits per 10 minutes per network so the service stays fast.
What the audit checks
AI search engines retrieve pages much like search crawlers do — then decide whether a passage is clear and trustworthy enough to quote. The audit covers both halves: access and answerability.
AI search readiness
- OAI-SearchBot, PerplexityBot, Claude-SearchBot access
- ChatGPT-User / Claude-User / Perplexity-User
- Live requests with each crawler user-agent
- Content visible without JavaScript
- nosnippet / max-snippet controls
- llms.txt and citable facts
Technical SEO
- noindex in meta and X-Robots-Tag
- Canonical tag and target
- Title, meta description, H1, headings
- HTTPS, HTTP→HTTPS, redirect chains
- Soft-404 test with a random URL
- Open Graph, hreflang, image alt, mixed content
Crawlability & speed
- robots.txt fetched and parsed (RFC 9309)
- Googlebot and Bingbot verdicts for this URL
- XML sitemap found and lists this URL
- Server response time (TTFB)
- gzip / Brotli compression
- HTML weight and internal links
Content answerability
- Answer-first opening paragraph
- Question-style headings
- Lists and tables engines can extract
- Content depth in the raw HTML
- Published / updated dates
- Author byline on articles
Entity & structured data
- JSON-LD syntax errors
- Required properties per Google type
- Organization / Person identity
- sameAs profiles and logo
- Schema that matches the page type
- Deprecated rich-result types
Why a normal SEO audit misses AI search
Most AI crawlers do not run JavaScript, follow their own robots.txt user-agents, and are often blocked by CDN “bot protection” that Googlebot gets through. A site can rank on Google and still be invisible to ChatGPT Search.
- Training crawlers (GPTBot, ClaudeBot, Google-Extended) are separate from search crawlers — block one without losing the other.
- Content rendered only in the browser is an empty page to OAI-SearchBot and PerplexityBot.
- Clear definitions, question headings and lists are what answer engines quote.
Search crawler vs training crawler
| Vendor | Search / citations | Training |
|---|---|---|
| OpenAI | OAI-SearchBot, ChatGPT-User | GPTBot |
| Googlebot | Google-Extended | |
| Anthropic | Claude-SearchBot, Claude-User | ClaudeBot |
| Perplexity | PerplexityBot, Perplexity-User | — |
| Apple | Applebot | Applebot-Extended |
Free tools for AI search visibility
AI crawler checker
See which AI bots robots.txt and your firewall let through.
robots.txt generator
Allow AI search, block training — with the right user-agents.
llms.txt generator
Create a clean llms.txt map of your key pages.
Schema validator
Paste JSON-LD and get required-property errors.
JSON-LD generators
Product, LocalBusiness, Article, FAQ and more.
AI crawler list
Every AI user-agent, what it does and how to control it.
Frequently asked questions
What does the SchemaCraft audit check?
Around 50 checks (the exact number depends on the page type) across five areas: technical SEO (indexability, canonical, titles, HTTPS, redirects, soft 404s), crawlability (robots.txt parsed per RFC 9309, XML sitemap, server response time, compression), content answerability (answer-first summary, headings, lists, dates, authorship), entity and structured data (JSON-LD syntax, required properties, Organization identity, sameAs) and AI search readiness (OAI-SearchBot, PerplexityBot and Claude access, JavaScript dependence, snippet controls, llms.txt). Every result shows the evidence it is based on.
Is the audit free?
Yes. The full report — every check, every fix and the AI crawler matrix — is free with no sign-up. To keep the service fast there is a limit of 20 audits per 10 minutes per network.
How do I check if ChatGPT can crawl my website?
ChatGPT Search uses the OAI-SearchBot crawler, and ChatGPT-User when a person asks about a specific page. Your robots.txt must not disallow them, your firewall or CDN must not block them, and the content must be in the HTML because these crawlers do not run JavaScript. The AI crawler checker tests all three in one run.
Should I block GPTBot?
GPTBot collects data for training OpenAI models; it is not the crawler behind ChatGPT Search results. You can block GPTBot and still appear in ChatGPT Search as long as OAI-SearchBot is allowed. The same split exists for Google-Extended vs Googlebot and ClaudeBot vs Claude-SearchBot.
What is the difference between SEO, GEO and AEO?
SEO makes pages rank in classic search results. AEO (answer engine optimization) makes content easy to lift as a direct answer — clear definitions, question headings, lists. GEO (generative engine optimization) makes a brand and its facts easy for AI systems like ChatGPT, Perplexity and Google AI Overviews to retrieve, trust and cite. All three start with the same foundation: crawlable, indexable, well-structured pages.
Does llms.txt help AI visibility?
llms.txt is an optional proposal that gives AI tools a Markdown map of your key pages. It is cheap to add, but no major AI search engine has confirmed using it for ranking, so the audit treats it as a low-weight check.
Why does the audit say it could not complete?
Usually because a firewall, CDN or bot-protection service blocked our server, the domain did not resolve, or the server timed out. No score is shown in that case because nothing was inspected. If your security setup blocks all automated requests, check that search and AI crawlers are allow-listed.
My score is high but I still do not rank. Why?
The score measures on-page readiness: whether crawlers can reach, read and understand the page. Rankings also depend on things a single-page audit cannot see — backlinks, brand authority, competition and how well the content matches search intent. A high score removes technical obstacles; it does not guarantee traffic.
Does valid Schema.org markup guarantee rich results?
No. Valid JSON-LD makes a page eligible; Google decides whether to show a rich result. Several types such as HowTo and most FAQ rich results are no longer shown, and markup must describe content that is visible on the page.