JSON Schema Generator Online - Generate Schema from JSON

JSON Schema is a vocabulary that allows you to annotate and validate JSON documents. It describes the structure, constraints, and data types of JSON data.

The tool supports draft-04, draft-07, and 2020-12. Draft-07 is recommended for most use cases.

Yes. The tool accepts any valid JSON and generates an appropriate schema. Complex nested objects and arrays are handled automatically.

The generated schema accurately represents the structure of your input JSON. However, since JSON alone cannot express constraints like minimum/maximum values or regex patterns, those will not be included.

The generated schema can be used with any JSON Schema validator library in your programming language of choice to validate JSON data against the expected structure.

If-then-else (draft-07+) allows conditional validation. If the schema in "if" validates, then "then" must also validate; otherwise "else" must validate. For example: if an object has a "price" field that must be a number, then require a "currency" field.

additionalProperties controls whether extra properties beyond those defined in "properties" are allowed. Set to false to reject unknown keys. In 2020-12, unevaluatedProperties provides even more flexible control over property validation.

JSON Schema describes JSON data structure for validation. OpenAPI (formerly Swagger) is a full API specification format that uses a subset of JSON Schema to describe request/response bodies. OpenAPI 3.0 uses JSON Schema draft-07 with some restrictions. OpenAPI 3.1 fully aligns with JSON Schema 2020-12.

draft-04 was the first widely adopted version (2013). draft-07 (2018) added if/then/else, $comment, and contentMediaType. 2020-12 is the latest with $defs replacing definitions, unevaluatedProperties, prefixItems for tuple validation, and improved vocabulary extensibility. draft-07 is the most widely supported for tooling.

The tool will display an error message indicating the JSON parse error. Please check your JSON syntax and try again.

Yes. Once you have visited the page, all functionality works offline thanks to local processing.

Yes. All processing happens locally in your browser. Your JSON data never leaves your device. No upload is needed.

oneOf requires exactly one of the listed sub-schemas to validate successfully, anyOf allows one or more sub-schemas to match, and allOf requires every sub-schema to pass simultaneously. For example, oneOf is ideal for discriminated unions like "paymentMethod" accepting either "credit_card" or "paypal", while allOf is used for schema composition and extending base definitions.

$ref allows referencing other schema definitions using JSON Pointer syntax, enabling DRY schema design. For example, "$ref": "#/$defs/address" points to a reusable address definition within the same schema. In draft-04 and draft-07, definitions go under "#/definitions/", while 2020-12 uses "#/$defs/" for this purpose. External $ref can also reference remote schema files.

JSON Schema alone cannot express cross-field constraints like "endDate must be after startDate" because each property is validated independently. Draft-07 introduced if-then-else for some conditional logic, but arbitrary cross-field comparisons require external validation in your application code. Use JSON Schema for type and structure validation, then add business rules programmatically.

In 2020-12, tuple validation uses the prefixItems keyword as an array of schemas where each position corresponds to an array element index. The items keyword (now unevaluatedItems in 2020-12) serves as a fallback for remaining elements beyond the tuple. This separation provides clearer semantics than draft-07 where items served dual purposes for both tuple and item validation.

JSON Type Definition (JTDF, RFC 8927) is a simpler, more restricted schema format designed specifically for HTTP APIs and protocol buffers. It supports fewer keywords than JSON Schema but is easier to implement and validate. JSON Schema is more expressive with conditional validation, pattern matching, and cross-references, while JTDF prioritizes simplicity and unambiguous type definitions.