The schema is too permissive — how do I add constraints?

Inferred schemas capture observed types but not business rules. Manually add constraints: string format (date-time, email), number minimum/maximum, string minLength/maxLength, enum values, and additionalProperties: false to reject unknown fields.

Should I infer from one sample or many?

Many. A single sample misses optional fields (fields absent in that instance become required) and union types (a field that is sometimes null). Provide multiple instances or manually mark fields optional after inference.

TASK

Infer a JSON Schema from sample data

Generate a JSON Schema that describes the structure of a JSON instance, for validation and documentation.

1

A JSON Schema describes the shape of JSON data — required fields, types, formats, constraints. Inferring a schema from sample data gives you a validation contract and documentation without hand-writing the schema.
2

JSON Schema Generator →
Paste a JSON instance (ideally with all fields populated). The tool infers types, required properties, array item types, and emits a JSON Schema (draft 2020-12).
3

JSON Validator →
Validate other instances against the generated schema to confirm they conform. Mismatches reveal fields that are missing, extra, or wrong-typed.

Frequently asked questions

The schema is too permissive — how do I add constraints?: Inferred schemas capture observed types but not business rules. Manually add constraints: string format (date-time, email), number minimum/maximum, string minLength/maxLength, enum values, and additionalProperties: false to reject unknown fields.
Should I infer from one sample or many?: Many. A single sample misses optional fields (fields absent in that instance become required) and union types (a field that is sometimes null). Provide multiple instances or manually mark fields optional after inference.