JSON-LD Validator
Validate JSON syntax and review JSON-LD context, type and structural issues locally, with errors and warnings kept separate.
Your JSON-LD
Paste JSON only, without script tags. Use an array or @graph for multiple objects. Up to 256 KiB.
Local syntax and selected structure checks. No remote context loading, full JSON-LD expansion, complete Schema.org validation or Google eligibility test.
Validation result
Understand your results
For developers and editors checking a JSON-LD block before embedding it in a page.
The validator first parses JSON with the existing numeric-precision safeguards. It then walks objects, arrays and graph entries to inspect selected JSON-LD keyword shapes. Syntax failures stop processing; malformed keyword values appear as structural errors, while missing context/type and empty values are review warnings. A warning can describe legitimate JSON-LD that needs context, not an invalid document.
How to use this tool
- Paste JSON without script tags. Wrap multiple documents in one array, or use @graph for related nodes.
- Run Validate JSON-LD and read structural errors separately from warnings and JSON parsing failures.
- Copy the indented JSON after reviewing it, then use a full JSON-LD processor and the relevant vocabulary/consumer validators where needed.
What the observations mean
JSON syntax, JSON-LD structural shape, vocabulary meaning and search eligibility are separate layers. The result distinguishes the first two locally; a parsed and formatted object is not a rich-result certificate.
Worked example
Check a graph with a nested author rather than assuming each nested node needs a new context.
{"@context":"https://schema.org","@graph":[{"@type":"Article","headline":"Café guide","author":{"@type":"Person","name":"Example Author"}}]}JSON syntax passes and Article/Person type labels are found. The graph inherits its context. A context-expansion limitation remains a warning even when the implemented shapes pass.
Limits and processing
Local subset check only: 256 KiB, depth 40, 4,000 visited values and 100 issue messages before a truncation notice. Unsafe numbers are rejected. Context URLs are not fetched; vocabulary ranges, aliases, advanced JSON-LD semantics and Google eligibility require separate tools.
Common problems and practical fixes
- Syntax parsing fails
- Remove script wrappers, trailing commas or comments; keep multiple objects inside a JSON array.
- Keyword values have the wrong shape
- Use string identifiers, string or string-array @type values, a valid context shape and object/array graph entries.
- Warnings remain after syntax passes
- Review context scope and empty facts. Use a full processor for alias expansion and vocabulary validation rather than suppressing warnings blindly.
Best practices
Retain the original block, correct the smallest issue first and validate the deployed page separately. Input is local and remote contexts are never resolved.
Practical tips
- Contexts inherit into nested nodes, not between sibling documents in a top-level array.
- An @id-only reference or graph container need not declare its own @type. Missing @type is deliberately not a universal error.
- No context URL is requested. A structural pass cannot prove that compact names expand to the intended vocabulary.
Frequently asked questions
Practical answers for using the JSON-LD Validator.
Does this validate all of Schema.org?
No. It checks JSON syntax and a documented subset of JSON-LD structure. It does not look up every type/property, enforce their ranges or expand contexts.
Why is missing @context a warning?
Full IRI keys can describe linked data without a compact-term context. Most Schema.org examples use a context, so its absence needs review rather than an automatic syntax rejection.
Can I paste several JSON objects?
Use one valid JSON array or @graph document. Adjacent objects separated only by whitespace are not a single JSON document and fail syntax parsing.