JSON minifier

A JSON minifier strips out every space, line break and indent, leaving the smallest text that still parses to the same data. {"a": 1} and {"a":1} are identical to any parser. For a pretty-printed file that is usually a 20 to 40 percent cut in raw bytes, depending on how deeply it was indented.

That number, on its own, is misleading. Almost every server sends JSON compressed, and compression is very good at repeated whitespace, so most of the indentation you just removed would have cost little on the wire anyway. This tool measures both: raw bytes, and the compressed size using whichever of gzip, deflate and Brotli your browser can run. Nothing is estimated.

Treat the compressed figure as a comparison, not a promise. Your host or CDN may use a different compression level or a different algorithm and will produce a different transfer size. What it tells you reliably is the shape of the saving – whether minifying moves the needle for your file, or whether it is noise.

Duplicate keys, and why your parser stays quiet

JSON allows the same key twice in one object. Parsers do not error. They keep the last value and discard the earlier one, silently:

{
  "id": 123,
  "name": "Priya",
  "id": 456
}

That parses cleanly and id is 456. The 123 is gone. When the duplicate comes from a bad merge or a template loop, the discarded value can be an entire nested object, and nothing reports it. This tool scans the raw text before parsing and names every duplicate: the key path, the line it first appeared on, and the line it repeated on. The output already reflects what the parser did, so fix the source if the earlier value mattered.

When the JSON will not parse

The error panel gives the line and column, prints the offending line with a caret under the exact character, and names the cause instead of repeating the browser message. It recognises trailing commas, single quotes, unquoted keys, Python None, True and False, NaN and Infinity, comments, and JSON that has been escaped into a string.

Where the fault is mechanical, a repair button appears. The repair is deterministic rather than a guess, and it lists every change afterwards. One change deserves a second look: NaN, Infinity and undefined have no JSON equivalent, so they become null. That loses information, which is why the tool tells you rather than doing it quietly.

Sorting keys

The sort option reorders keys alphabetically, all the way down. The data is unchanged and the JSON stays valid, but key order is not preserved. Useful when you want two files to diff cleanly or a stable output to commit. Leave it off if the order in your file carries meaning.

When this is not the right tool

For reading a large, deeply nested payload, a textarea is the wrong shape – a collapsible tree view like JSON Editor Online is built for that. If minifying belongs in a build step, do it locally with jq -c . or a bundler plugin rather than a browser tab. For converting JSON into CSV, YAML or XML, CodeBeautify carries a far broader converter set than we do today. And JSON5 or JSONC config files are not JSON: this tool will offer to strip the comments and trailing commas, but if you need them preserved, use a JSON5-aware editor.

Everything here runs in your browser. There is no upload, no account and no file-size cap beyond your device’s memory. Large files are processed on a background thread so the page stays responsive.