JSON formatter: pretty, minified, and actually valid

Pretty-print: for humans and pull requests

Indented JSON is reviewable JSON. Two spaces is the convention in most JavaScript codebases and matches the default in `JSON.stringify(value, null, 2)`; four spaces suits editors configured that way. The formatting choice matters less than consistency, because mixed indentation turns every later diff into noise. If your JSON is generated, set the indentation in the generator once and let the formatter enforce it locally for pasted payloads.

Minify: for payloads on the wire

Whitespace in JSON is optional, so stripping it is free savings. On a 200 KB API response the difference is typically 15–30 percent of bytes, which matters when the budget is a mobile connection or a cache header you are trying to respect. Minifying is also the right move before storing a payload in a database column or embedding it in a URL — just minify as a build step rather than by hand, since hand-minified JSON is where debugging goes to die.

Validation: the errors you will actually hit

JSON is a deliberately small grammar, which is why its errors are so repetitive. Almost every invalid JSON error traces back to one of five habits picked up from JavaScript or YAML:

• Trailing commas — `{"a": 1,}` is valid JavaScript and invalid JSON.
• Single quotes or unquoted keys — `{a: 'b'}` must become `{"a": "b"}`.
• Comments — `// note` and `/* note */` have no place in JSON at all.
• Unescaped control characters — literal newlines inside a string must be `\n`.
• Concatenated objects — two top-level objects side by side, usually from a log file with one JSON document per line.

A formatter should tell you the line, the column and show the excerpt. “Unexpected token” alone leaves you scrolling; “line 84, column 17 — trailing comma before }” is a fix.

Why validate before you debug

When an integration fails, the instinct is to inspect code. More often the payload itself is malformed: a template emitted a stray comma, a log pipeline truncated a line, or a copy-paste from a document introduced smart quotes. Running the JSON formatter first separates “the data is wrong” from “the code is wrong” in seconds, and the exact line number usually points straight at the template that produced it.

Large files and privacy

API responses and database exports frequently contain tokens, personal fields or customer data. A server-side JSON formatter means uploading that payload to a third party to have it handed back — a data-protection question with no upside, since formatting is a trivial local operation. The YamlBox JSON formatter parses in the browser with no network calls, so it works offline and keeps your payload on your machine; the same page gives you indentation choices, key sorting for stable diffs, and live statistics such as depth and key count.