Free schema validator: find broken JSON-LD and the fields each type is missing.
Updated
This schema validator reads every JSON-LD block on one page, tells you which ones fail to parse, and checks each item against the properties Google lists as required and recommended for its type. It also reads the page after JavaScript runs, so schema added by a plugin or a tag manager shows up too.
Use it after a theme change, a new SEO plugin or before you ship a template. Results show in full on the page, with no signup and no email.
- Free
- No signup
- Results in about 20 seconds
- ITEMS
- 3
- ERRORS
- 1
- WARNINGS
- 1
- BlogPostingWARNINGS
- BreadcrumbListERRORS
- FAQPageVALID
What the schema validator checks on your page
Each line is a test our code runs on the page you enter. Nothing is guessed by a model.
Every JSON-LD block
We find each application/ld+json script, parse it, and point to the line where a broken one stops.
Served and rendered
We read the HTML your server sends and the page after JavaScript runs, and tell you which types appear only after.
The schema.org context
Each block must name schema.org, or its types are just words to a parser.
Required properties by type
Organization, LocalBusiness, Article, Product, FAQPage, BreadcrumbList, WebSite and HowTo, against Google's own lists.
Values in the right form
Dates in ISO 8601, full https addresses for url, logo, image and sameAs, and prices as plain numbers.
The schema.org vocabulary
A crawler side validation flags types and properties that don't exist in schema.org, across JSON-LD, Microdata and RDFa.
Who runs the site
Organization schema on your homepage with a logo and sameAs profiles, the same test Site Health runs.
Retired rich results
FAQPage and HowTo are flagged as types Google no longer shows, so you don't chase a result that can't appear.
From a page address to every broken field in five steps.
- 01
Paste a page
A product page, a blog post or your homepage.
- 02
We fetch it twice
Once as your server sends it, once with JavaScript run.
- 03
We parse each block
Broken JSON is caught and pointed to by line and column.
- 04
We check each item
Against Google's required and recommended properties for its type.
- 05
You fix and recheck
Open the item, add what's missing, and run it again.
Why valid schema helps search engines and AI answers
Structured data is a label that says what a page is: a product at this price, an article by this author on this date, a business at this address. Google uses it to decide which search features a page can have, and only when the required properties are there.
For SEO, a broken block is worse than none, because the whole block is thrown away and nothing tells you. For AEO, answer engines lean on pages whose facts are easy to read, and schema states those facts plainly. Google's own guide to AI features says no special schema is needed for AI Overviews, and asks that the markup you have matches the visible text.
For GEO, the gap is JavaScript. Many sites add JSON-LD through a tag manager or a script. Google renders the page and finds it; the crawlers behind ChatGPT and other assistants mostly read the HTML as served. That is why this validator reads both versions and tells you which one carries your schema.
Errors, warnings and notices in your schema report
An error stops an item from working: a block that doesn't parse, a block with no schema.org context, or a required property left out. Fix these first.
A warning means the item works but something weakens it: a date in the wrong form, schema that only appears after JavaScript, a missing Organization on the homepage. Recommended properties and retired types are notices, shown so you can decide.
Valid here means valid against the rules we check, not a promise of a rich result. Google decides what to show, and its Rich Results Test is the last word on its own features.
| VERDICT | WHAT IT MEANS |
|---|---|
| Errors to fix | Something is broken. At least one block doesn't parse or one item misses a required property. |
| With warnings | It works, with gaps. Every item parses and has its required fields, but a check found something to tidy. |
| Valid | Nothing to fix. Every block parses and every item we check has what it needs. |
| No schema | Nothing to read. The page carries no JSON-LD, served or rendered. |
The schema fixes that come up most
A block won't parse: find the stray comma.
Open the block, go to the line we point at, and look for a trailing comma, an unescaped quote or a comment. JSON allows none of them.
A required property is missing: add the real value.
Use the value the page shows. A Product needs a name and an offer with a price; a LocalBusiness needs a name and an address.
Schema loads with JavaScript: move it into the HTML.
Print the JSON-LD in the server template or through your CMS's head field, so crawlers that run no scripts read it too.
No Organization on the homepage: add one block.
Name, url, logo and sameAs links to your profiles. It is the block engines use to tie your pages to your brand.
A minimal Organization block for your homepage, with the four properties our check looks for:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Northwind",
"url": "https://northwind-crm.com",
"logo": "https://northwind-crm.com/logo.png",
"sameAs": [
"https://www.linkedin.com/company/northwind"
]
}
</script>When to validate a page's schema.
- After you switch SEO plugins: two plugins often print two Organization blocks, or the new one drops fields the old one filled.
- Before a template ships: check one product or post on staging so every page built from it starts clean.
- When rich results vanish: see whether a block broke, a required field went missing or the type is one Google retired.
- When schema comes from a tag manager: confirm whether it is in the served HTML or only appears after scripts run.
- On a client audit: agencies validate a prospect's homepage, a product page and a post, and lead with what is broken.
One page is a spot check. Zivooo checks every page.
Site Health in Zivooo validates the structured data on every page of your site each month, on your own keys, and every article Zivooo writes carries its Article and FAQ schema from the start. $10 a domain a month after 7 days free, no card.
- Structured data errors found across the whole site
- Organization and breadcrumb checks on every audit
- Articles published with their schema already in place
- Whether ChatGPT, Gemini and Google name your brand
This schema markup validator vs Google's Rich Results Test
Google retired its Structured Data Testing Tool and split it in two: the Rich Results Test, which says whether a page can get Google's search features, and the Schema Markup Validator at validator.schema.org, which checks any schema.org markup without Google's rules. Both are free and both are the reference. This structured data testing tool sits between them: it checks Google's required properties for the common business types, reads the page as served and rendered side by side, and adds the homepage Organization check, which neither of them runs. It doesn't preview rich results, and it doesn't replace Google's test.
Schema validation, answered.
What is a schema validator?
A tool that reads the structured data on a page and tells you whether it is written correctly: whether the JSON parses, whether the types and properties exist, and whether each item has the properties a search engine needs to use it.
How do I check if my website has schema?
Enter a page above. We list every JSON-LD block we find, served or added by JavaScript. You can also view the page source and search for application/ld+json, though that misses schema added by scripts.
What's the difference between the Rich Results Test and the Schema Markup Validator?
The Rich Results Test checks Google's rules and says which Google features a page can get. The Schema Markup Validator checks schema.org markup in general, with no Google rules. This validator checks Google's required properties for common types and reads both versions of the page.
Does schema markup help with AI search?
It states your facts in a form any parser can read, which helps answer engines get them right. Google says AI Overviews need no special schema and that markup should match the visible text, so treat it as clarity, not a trick.
Why does my schema show in Google's test but not in yours?
Google renders JavaScript, and so do we on the second read. If we report the schema as missing from the served HTML, it was added by a script: Google sees it, many AI crawlers don't.
Do you check Microdata and RDFa?
We count them and the schema.org vocabulary check covers them, but the item by item checks read JSON-LD only. For a Microdata page, the Schema Markup Validator shows each item.
Write the schema you're missing.
- SCHEMA & AUTHORITYSEOAEOSchema Markup GeneratorBuild JSON-LD for eight schema types from a short form.Generate schema
- SCHEMA & AUTHORITYSEOJSON-LD GeneratorCombine schema types into one linked @graph block.Generate JSON-LD
- SCHEMA & AUTHORITYAEOSEOFAQ Schema GeneratorTurn your questions and answers into FAQPage JSON-LD.Generate FAQ schema
- AI READINESS & TECHNICALAEOGEOSEOAI Readiness CheckerSee if AI search bots can reach, read and quote a page.Check my AI readiness
Keep every page's schema valid.
Your keys, your costs, $10 a domain. 7 days free, no card.