Free tool

Schema Validator

Enter a URL or paste JSON-LD. The validator finds every structured data block and checks it for syntax errors, missing required properties, bad dates and relative URLs. It also says which Google rich results each item may qualify for. Free, no signup.

Or paste JSON-LD, or a page's HTML, below.

What it checks

What does a schema validator check?

GetIntel's schema validator is a free tool that checks the structured data on a page. Schema markup, usually JSON-LD, tells search engines what a page describes: an article and its author, a product and its price, a company and its logo. The validator reads it the way a parser does and flags what breaks it. For the types Google turns into rich results, it also checks the properties Google requires and recommends.

Does schema markup help AI citations?

No, not on our data. Among the 40 pages ChatGPT, Perplexity, Gemini and Google AI Overviews cited most when answering 130 buyer questions about AI visibility tools, checked by GetIntel on 11 August 2026, the 29 pages carrying JSON-LD averaged 118.3 citations. The 11 without it averaged 121.3. That is a small sample in one category, but it shows no lift. In a separate GetIntel test of 615 cited pages in August 2026, FAQ schema looked like it helped until each publisher was counted once, and then the gap disappeared. Both studies are in our llms.txt and schema test and our FAQ schema study.

So keep your schema correct because it is cheap and it earns Google rich results: star ratings, prices, breadcrumbs, logos. Don't add it expecting AI engines to cite you more. What our data points to instead is being on the pages they already cite, which is what GetIntel's AI visibility tracking shows you.

What errors did the validator find on real sites?

Before launch, on 18 October 2026, we ran the validator on 10 sites: SaaS homepages, a pricing page, a blog, a store's product page, a recipe and a news homepage. Three had no JSON-LD in the HTML their server returned. Of the 7 that did, 3 had at least one error. One article gave its dates as "February 15, 2023" rather than ISO 8601, the format Google's documentation asks for. One product page kept its star rating in a separate block with no @context or @type, so parsers may ignore it. One software homepage had no rating, which Google needs before it shows the software rich result. Errors like these are silent: the page looks fine, and the markup may be ignored.

What can't this schema validator check?

It reads the HTML your server returns and does not run JavaScript, so schema added by a tag manager won't show up. Paste the rendered code instead. It covers JSON-LD, not microdata or RDFa, though it tells you when a page uses them. And it checks the properties Google documents for each type, not the whole schema.org vocabulary. Google's own Rich Results Test is the final word before you ship.

FAQ

Frequently asked questions

GetIntel's schema validator finds every JSON-LD block on a page or in pasted code. Each block must parse as JSON and declare a schema.org @context and an @type, and it checks dates and URLs on each item. Where Google defines required properties for a rich result, a missing one is an error, and a missing recommended one is a warning. It knows the rules for Article, Product, ProductGroup, Organization, Breadcrumb, FAQ, HowTo, LocalBusiness, SoftwareApplication, Event, Recipe, Video, JobPosting, Review and WebSite.

Not on our data, which is a small sample in one category. Among the 40 pages ChatGPT, Perplexity, Gemini and Google AI Overviews cited most when answering 130 buyer questions about AI visibility tools (GetIntel, checked 11 August 2026), the 29 with JSON-LD averaged 118.3 citations and the 11 without averaged 121.3. A separate GetIntel test of 615 cited pages (August 2026) found that FAQ schema's apparent edge disappeared once each publisher was counted once. Schema is worth keeping correct for Google's rich results, not as an AI citation lever.

GetIntel's schema validator reads the HTML your server returns and does not run JavaScript. If your structured data is added by a tag manager or a client-side script, it won't appear here, while Google, which renders pages, may still see it. Paste the rendered code from your browser's developer tools to check it, or move the JSON-LD into the server-rendered HTML.

For Google rich results, mostly no, and GetIntel's schema validator says so when it finds FAQPage markup. Since August 2023 Google shows FAQ rich results only for well-known government and health sites, and it stopped showing HowTo rich results the same year. Valid FAQPage markup does no harm, and other engines can read it. Most sites still will not get the expandable result in Google.

Google's test renders the page and is the final word on what Google will show. This validator is faster for checking raw JSON-LD, accepts pasted code without a URL, lists every block on the page together, and says plainly which types no longer earn a rich result. Use this one while writing the markup, and Google's test before you ship.
Your next move

Valid schema can earn Google rich results. On our sample, it showed no link to AI citations.

GetIntel shows which pages AI engines cite instead of yours, and turns each gap into a ranked task.

First full scan free · No card · Results in about 3 minutes

More Free Tools

Explore other tools to boost your productivity

Built for Your Founder Type

See how GetIntel fits founders, teams, agencies, and indie hackers

Correct schema is housekeeping. Being cited is the goal.

GetIntel tracks whether AI engines cite you across ChatGPT, Perplexity, Gemini and Google AI Overviews daily (plus Claude on your own Anthropic key), and turns every gap into a ranked task. From $29/mo. First scan free, no card.