What JSON is for

JSON is a text format for representing structured data. It was designed to be easy for humans to read and easy for machines to parse. Although it grew out of JavaScript, it is completely language-independent: virtually every programming language can read and write it.

Its main job is data interchange — moving structured information between systems. When a web app asks a server for your profile, the server typically replies with JSON. When a tool stores settings, it often saves them as JSON.

The building blocks

JSON is built from just a few value types, which can be nested inside one another:

• Objects — an unordered set of key/value pairs wrapped in curly braces { }. Keys are always strings in double quotes.
• Arrays — an ordered list of values wrapped in square brackets [ ].
• Strings — text wrapped in double quotes, e.g. "hello".
• Numbers — integers or decimals, written without quotes, e.g. 42 or 3.14.
• Booleans — the literal values true or false.
• null — represents the intentional absence of a value.

A worked example

Here is a small JSON document that uses every value type:

{
  "name": "Ada Lovelace",
  "age": 36,
  "active": true,
  "roles": ["admin", "author"],
  "address": {
    "city": "London",
    "postcode": null
  }
}

Reading it: the top-level value is an object with five keys. "roles" is an array of two strings, and "address" is a nested object. "postcode" is null, meaning no value is set.

Rules that trip people up

• Keys and string values must use double quotes — single quotes are invalid.
• No trailing commas: the last item in an object or array must not be followed by a comma.
• Comments are not allowed in standard JSON.
• The whole document must have exactly one top-level value (usually an object or array).

If you hit a parsing error, paste your JSON into the formatter above and it will point to the exact line and reason.