JSON Formatter & Validator

How to use

1. Paste or type JSON into the input pane at the top. The counter beneath it tracks character count and byte size as you edit.
2. Set the Indentation dropdown to 2 or 4 spaces. This choice only affects Format.
3. Tick Sort keys A→Z (recursive) if you want object members reordered at every nesting level. Arrays keep their original order.
4. Press Format to pretty-print, Minify to strip whitespace, or Validate to check syntax without producing output. Ctrl/Cmd+Enter is a shortcut for Format.
5. Read the Object keys, Max depth and Size stats, then Copy the result or Download it as formatted.json or minified.json.

How it works

All three buttons run the same short pipeline. Your text is handed to the browser’s native JSON.parse, which either returns a JavaScript value or throws at the first character that breaks the RFC 8259 grammar. When it throws, the tool reads the line, column or byte offset out of the error message, selects that line in the input, and scrolls to it. When it succeeds, the value is optionally passed through a recursive key sort, then serialised again with JSON.stringify — using your indent for Format, no spacing for Minify, and skipped entirely for Validate, which stops the moment the parse succeeds.

Take this input with Sort keys on and Format set to 2 spaces:

{"sku":"TT-90","tags":["b","a"],"stock":{"qty":12,"bin":"D4"}}

The parse yields an object. The recursive sort reorders the top-level keys to sku, stock, tags, and reorders stock’s members to bin, qty — but leaves the tags array as ["b","a"], because element order there is meaningful. Serialising with two-space indent gives:

{
  "sku": "TT-90",
  "stock": {
    "bin": "D4",
    "qty": 12
  },
  "tags": [
    "b",
    "a"
  ]
}

The stats reflect the parsed value, not the text: Object keys reads 5 (three at the top level plus two inside stock), Max depth reads 2 (the stock object and the tags array both sit one level down), and Size is the UTF-8 byte length of the output. The walk that computes depth is iterative, so a file nested tens of thousands of levels deep reports its stats instead of overflowing the call stack.

Use cases & limitations

A minified webhook payload printed on one endless line is unreadable; Format turns it back into something you can scan and diff. Reach for this when you are debugging an API response, tidying a hand-edited config file, normalising a test fixture so its key order is stable, or shrinking a document before you paste it somewhere size-constrained. Pairing it with Text Diff is a common move: format two responses the same way, then compare them line by line so only the real changes stand out. If the JSON is a token payload rather than plain data, the JWT Decoder splits and decodes it for you.

The honest limitation is that this is a strict parser, not a repair tool. It will not guess your intent, close a missing bracket, or strip a stray trailing comma — it reports the first grammar error and stops, so genuinely broken input is fixed one error at a time by hand. It formats valid JSON; it does not rescue invalid JSON.

Privacy note

JSON routinely carries access tokens, personal details and internal identifiers, so where it goes matters. Every step here — parse, sort, serialise, byte counting — runs through native browser functions inside your tab. No request is made at any point, nothing is stored, and the download is generated locally from the text already on screen. Open your browser’s network tab and format the most sensitive document you have: it stays empty.