Free tool

FAQ schema checker: test or generate FAQPage JSON-LD

Paste a URL to check your existing FAQ schema for errors, or generate valid JSON-LD from the questions already on the page. Includes what actually changed when Google removed FAQ rich results on 7 May 2026.

Free · no account · under 3 seconds. One fetch, no LLM call.

What this FAQ schema tool checks

  • Existing FAQPage JSON-LDReads @graph-nested and mainEntity/hasPart schema, not just a top-level FAQPage block — most naive checkers miss both.
  • Missing acceptedAnswerFlags each Question node with no answer, which is invalid and the most common real-world error.
  • Question-shaped headingsH2–H4 headings ending in "?" plus the paragraph that follows.
  • Accordion FAQs<details>/<summary> patterns, the common no-JavaScript accordion.
  • Definition-list FAQs<dl>/<dt>/<dd> markup — older than schema.org, still common in hand-built pages.

FAQ schema questions

Does FAQ schema still work in 2026?

The schema.org FAQPage type is still valid and causes no problems to have on your page. What changed is what it earns you: Google removed FAQ rich results from Search entirely on 7 May 2026, extending a restriction it first applied in August 2023 (limiting FAQ snippets to authoritative government and health sites). As of the May 2026 change, no site gets a FAQ rich result, regardless of schema.

Should I remove my existing FAQ schema?

No need to. Google’s own documentation confirms unused structured data does not affect Search. Leaving it in place is harmless; it simply no longer earns a rich result.

Is there any reason to add FAQ schema now?

A narrower one than "get a Google snippet." Explicit Question/Answer structure removes ambiguity about where a question ends and its answer begins, for anything else parsing the page — a screen reader, a browser extension, or an AI crawler extracting content. None of those benefit from Google’s former rich-result carousel, but all of them still read the markup.

How does the generator decide what to put in the schema?

It only uses text already published on your page — question-shaped headings with the paragraph that follows, <details>/<summary> accordions, and <dl>/<dt>/<dd> definition lists. It does not write new questions or answers. If your page has no visible FAQ content, there is nothing to generate from and it says so.

What counts as an FAQ schema "issue"?

The most common one: a Question node with no acceptedAnswer, which is invalid per schema.org and gets flagged individually so you can see exactly which question is broken.

Valid markup does not mean complete answers.

FAQLens tests every buyer-critical question against what an AI can actually extract from your page, and drafts the wording for what is missing.