JSON Schema Generator & Validator

What JSON Schema is for

A JSON Schema describes what a valid document looks like: which fields exist, their types, which ones are required and what values are allowed. Teams use it to document APIs (OpenAPI uses it for request and response bodies), to validate input on a server, and to catch bad test data before it reaches production.

Generating a schema from an example

Paste a real response or config file and the generator writes a draft 2020-12 schema:

• Every object gets its properties, with a type for each one. Whole numbers become integer; numbers with decimals become number.
• In arrays, all items are merged. If a field appears in some items but not others, it is kept but not required. If a value is sometimes null, the type allows null too.
• Strings that look like ISO dates, email addresses, http(s) URLs or UUIDs get a format.
• Turn off Allow extra properties to add "additionalProperties": false, which rejects any field the schema doesn't list.

A generated schema is a starting point. Tighten it by hand, for example with minimum, maxLength, enum or pattern, before relying on it.

Validating JSON

Switch to Validate JSON, paste a schema and a document, and every problem is listed with the path of the value, a plain explanation and the rule that failed, for example $.orders[0].qty: "0 is less than the minimum 1".

Supported keywords: type, enum, const, properties, required, additionalProperties, patternProperties, propertyNames, minProperties/maxProperties, dependentRequired, items, prefixItems, additionalItems, contains, minItems/maxItems, uniqueItems, minLength/maxLength, pattern, format, minimum/maximum and their exclusive forms, multipleOf, allOf, anyOf, oneOf, not, if/then/else and $ref within the same schema.

For production use, validate in your application with a full implementation such as Ajv (JavaScript), networknt json-schema-validator (Java) or jsonschema (Python). This page is for quick checks while you work.

Frequently asked questions