Validate JSON Against a JSON Schema

JSON Schema describes the shape of a JSON document — required keys, allowed types, value ranges, regex patterns. Validating against a schema catches contract drift between API producers and consumers, broken fixtures, and invalid configs before they hit production.

We target the Draft-07 essentials most schemas in the wild rely on: type, required, properties, additionalProperties, items (including tuples), enum, const, minimum/maximum (and exclusive variants), minLength/maxLength, minItems/maxItems, uniqueItems, pattern, and format.

Local refs only. References like #/$defs/User or #/definitions/User resolve at validation time. External refs (URLs, other files) are not fetched — the validator silently skips them so a stray external $ref will not fail your check.

No. This is a pragmatic subset for the 90% case — fast, dependency-free, and runs entirely in the browser. oneOf/anyOf/allOf/not, if/then/else, dependencies and patternProperties are not yet implemented. For full Draft 2020-12 conformance, use Ajv in your own pipeline.

Never. Validation runs entirely client-side. There are no servers, no logs of your data, and no behavioural analytics on your inputs — your JSON and schema stay in your tab.

Each violation includes the JSONPath of the offending value, the rule that failed, and the expected vs actual value. Click any error to jump to the line in the JSON editor — making this a practical JSON checker for fixing real-world data.