MySQL JSON_TYPE(): Return Values, NULL, and Examples

MySQL JSON_TYPE() returns a utf8mb4 string describing the type of a JSON value.

Syntax

JSON_TYPE(json_value)

Return Values

For a valid non-NULL argument, JSON_TYPE() can return:

Basic Examples

SELECT
  JSON_TYPE('{"name": "Ada"}') AS object_type,
  JSON_TYPE('[1, 2, 3]') AS array_type,
  JSON_TYPE('"hello"') AS string_type,
  JSON_TYPE('true') AS boolean_type;

The result types are OBJECT, ARRAY, STRING, and BOOLEAN.

Number Types

SELECT
  JSON_TYPE('1') AS integer_type,
  JSON_TYPE('1.25') AS number_type;

Typical output is INTEGER and DOUBLE. MySQL can also report DECIMAL for exact-decimal scalar values.

SQL NULL vs JSON null

These two cases are different:

SELECT
  JSON_TYPE(NULL) AS sql_null,
  JSON_TYPE('null') AS json_null;

• JSON_TYPE(NULL) returns SQL NULL.
• JSON_TYPE('null') returns the string 'NULL', meaning the JSON null literal.

Inspecting a Nested Value

Use JSON_EXTRACT() with JSON_TYPE():

SET @doc = '{"user": {"id": 10, "roles": ["admin"]}}';

SELECT
  JSON_TYPE(JSON_EXTRACT(@doc, '$.user')) AS user_type,
  JSON_TYPE(JSON_EXTRACT(@doc, '$.user.id')) AS id_type,
  JSON_TYPE(JSON_EXTRACT(@doc, '$.user.roles')) AS roles_type;

This returns OBJECT, INTEGER, and ARRAY.

Invalid JSON

JSON_TYPE() expects a valid JSON value. Invalid JSON produces an error:

SELECT JSON_TYPE('hello');

The text hello is not a valid JSON string literal because JSON strings require quotes. Use:

SELECT JSON_TYPE('"hello"');

If you need to validate text first, use JSON_VALID().

Summary

• JSON_TYPE() returns a string describing a JSON value’s type.
• SQL NULL and JSON null are different cases.
• Numeric values may be reported as INTEGER, DOUBLE, or DECIMAL.
• Invalid JSON raises an error.
• Combine JSON_TYPE() with JSON_EXTRACT() to inspect nested values.

See the MySQL 8.4 JSON attribute functions.