JSON Schema Types

Text — any sequence of characters: letters, digits, symbols, spaces. The most versatile type; use it for names, descriptions, addresses, codes, and any other free-form text. Supports formats for semantic validation.

Any numeric value — whole or decimal. Covers integers, fractions, and scientific notation. Use when a fractional value may be needed: prices, distances, weights, ratings. For whole numbers only, prefer integer.

A whole number — no fractional part. Use when only round values make sense: counts, IDs, ages, page numbers, quantities. If fractions are possible, use number instead. Decimal notation like 1.0 is accepted as long as the fractional part is zero.

A true/false value. Use for on/off switches, yes/no answers, active/inactive flags, or any field that has exactly two states. Note: JSON booleans are lowercase true and false, not strings or numbers.

The intentional absence of a value — "this field has no value." In JSON, null is its own distinct type, not zero, not an empty string, not false. Often combined with another type (e.g. string or null) using anyOf to mark a field as optionally empty.

An ordered list of values. Items appear in sequence and can be accessed by position (first, second, third…). Use for collections where order matters: tags, search results, cart items, a list of phone numbers. The optional itemskeyword constrains what type each element must be.

A structured set of named fields (key-value pairs). Use for records, profiles, addresses, and settings — any entity with multiple attributes. The propertieskeyword defines the shape of the object, and requiredlists which fields must always be present.