Schema markup generator

Structured data doesn't change your words. It labels them so a machine knows what they mean. Its documented payoff is in Google Search: rich results for the types Google supports. For AI citation the evidence is thin. An Ahrefs observational test of 1,885 pages found citations barely moved after schema was added, live-fetch tests found AI systems reading visible text rather than JSON-LD, and Google says no special schema is needed for AI Overviews or AI Mode. The evidence scoreboard grades it Open, and whether schema causes citation is a question this site has pre-registered rather than assumed.

What each type is good for on Google today: Article helps Google understand headline, author and dates. Person, with a sameAs list of your profiles, describes the author; whether that author signal moves AI citations is being tested, not claimed. FAQPage is valid markup, but since August 2023 Google shows FAQ rich results only for well-known, authoritative government and health sites. HowTo rich results were removed from Google in 2023. Both still describe your content to any parser that reads schema.org.

1. Pick the type that matches the page. A blog post is an Article. An about or author page is a Person. Use FAQPage only if the page shows a list of questions with their answers, and HowTo only if it walks through numbered steps.
2. Fill the fields. The output updates as you type. Leave a field blank and it is dropped from the code, so you never ship an empty "image": "".
3. Copy the JSON-LD. Press Copy. The code is a plain object starting with "@context": "https://schema.org".
4. Wrap it in a script tag. Paste it between <script type="application/ld+json"> and </script>, in the head or body of that one page. Most CMSs have a field or plugin for custom head code.
5. Test before you publish. Run the live URL or the code through Google's Rich Results Test for Google features, and the Schema Markup Validator for general schema.org correctness.

The FAQPage tab takes one question and answer per line, split by a pipe. Type this:

What is llms.txt? | A Markdown file at /llms.txt that maps your key pages for AI models.
Does it work yet? | It's an early proposal.

and the generator returns:

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is llms.txt?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "A Markdown file at /llms.txt that maps your key pages for AI models."
      }
    },
    {
      "@type": "Question",
      "name": "Does it work yet?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "It's an early proposal."
      }
    }
  ]
}

Only the first pipe splits the line, so an answer can contain its own |. Lines without a pipe are skipped. The HowTo tab works the same way, with Step name | instruction, and numbers the steps for you. For a test of whether markup like this changes AI citations, see the schema study; if you are setting up an llms.txt file at the same time, the llms.txt generator builds one.

It does four types with the fields most pages need. It removes empty fields, and it escapes < as \u003c so a stray closing script tag in your text can't break the page. That escape is valid JSON, and parsers read it as the normal character.

It doesn't:

• Validate your dates or URLs. Type 2nd July and it goes into the code as-is. Use ISO 8601 dates such as 2026-07-02, and full absolute URLs starting with https://.
• Cover every type. There's no Product, Organization, LocalBusiness, BreadcrumbList or Event tab. Breadcrumbs and Organization are usually better generated sitewide by your theme or CMS.
• Link entities with @id. The Article author is a plain Person with a name. If you want the author to point to your full Person block, add matching @id values by hand.
• Check the page itself. Google's guidelines require markup to describe content a visitor can see. The tool can't tell whether your FAQ answers are actually on the page.

Schema is one part of a technical check. The citation readiness checker looks at a page more broadly, and the robots.txt tester confirms crawlers can reach it at all.