JSON Formatter

How to use

1. Paste JSON into the left box or click "Open file" to load a .json file. Syntax is checked as you type, and the tree view on the right updates live.
2. Pick an indent (2 spaces, 4 spaces or tab) and click "Format" for readable multi-line output, or "Minify" to squeeze it onto one line. Tick "Sort keys" to order every object's keys alphabetically.
3. If there's a syntax error, you'll see the exact line and column. Click "Show error" to select the offending character, then use "Copy" when you're done.

Worked example

Trailing comma: entering the four lines {, "name": "Pockez",, "tags": ["json", "tool"],, and } reports "Line 4, column 1: there's a comma after the last item". The error points at the closing }, but the comma to delete is at the end of line 3.

Big integers stay exact: for {"orderId":9007199254740993,"price":1.50}, JavaScript's JSON.parse returns 9007199254740992 and a price of 1.5. After formatting here you still get 9007199254740993 and 1.50, plus a note that 1 integer is beyond the safe range.

Sorting keys: format {"name":"Alice","age":30,"tags":["a","b"]} with "Sort keys" ticked and a 2-space indent, and the keys come out as age, name, tags, while the a and b inside the tags array keep their order.

Reading the results

• "Valid JSON" means the whole document follows the JSON standard (RFC 8259), so any programming language can read it.
• The line and column in an error mark the first spot the parser couldn't understand. The real fix is often just before it: a missing comma, for example, is reported at the start of the next key.
• In the tree view, { 3 } means an object with 3 fields and [ 2 ] means an array with 2 items. Numbers with a dotted underline are big integers; hover for details.
• A big-integer note doesn't mean your data is wrong. It means a receiver that reads numbers as double-precision floats, like JavaScript, will lose precision, so consider sending those values as strings or reading them with BigInt.

How it works

• A purpose-built parser reads the text character by character using the RFC 8259 and ECMA-404 grammar and stops at the first error. It doesn't rely on the browser's JSON.parse, so error positions and messages are the same in every browser.
• Formatting and minifying only rearrange whitespace. Strings keep their original escapes (\u00e9 stays \u00e9), numbers keep their original spelling (1.50 and 1e400 are untouched), and key order and duplicate keys are preserved.
• Sort keys orders every object recursively by UTF-16 code unit, the same as JavaScript's default sort. Duplicate keys keep their relative order, and arrays are never reordered.
• Integers whose absolute value exceeds 2^53−1 (9007199254740991) are flagged as big integers. Numbers with a decimal point or exponent are not flagged.

Limitations

• Only strict JSON is accepted. Comments, trailing commas, single quotes, NaN, undefined and other JSON5 or JavaScript syntax are reported as errors.
• Only the first error is reported at a time; fix it and the check moves on to the next one.
• Nesting deeper than 1000 levels stops parsing with a message instead of freezing the browser.
• Files of tens of megabytes can slow the browser down. The tree view only opens the first two levels; "Expand all" builds every node, so for big files open just the parts you need.

FAQ

💬 Can't find an answer? Report a problem with this tool or send a suggestion