JSON Formatter & Validator

Formatting is the easy half. The useful half is telling you precisely where a document stops being JSON, and what it contains once it parses.

Strict, deliberately

Trailing commas, unquoted keys, single quotes and comments are rejected. They are JavaScript object syntax or JSON5, not JSON, and a formatter that repairs them hands you back something the source never said — which matters when the source is evidence rather than a file you are editing.

Errors are located here, not read off the engine

V8 stopped putting an offset in its JSON error messages, so a tool built on parsing that message reports line 1 for every error — useless on a 400-line document. This page scans the text against the grammar in RFC 8259 and reports the line, the column and the line itself. The scan runs only after the native parse has already failed, so the ordinary path is still one fast parse.

Duplicate keys

A key repeated inside one object is legal JSON. JSON.parse keeps the last and discards the rest in silence, so {"a":1,"a":2} becomes {a:2} and the 1 is simply gone. That is also how a value gets smuggled past something that read the first occurrence — so the raw text is scanned for it, because the parsed object no longer holds the evidence.

Nothing is executed

JSON.parse is a data parser, not an evaluator, and this page never calls eval or new Function. A payload crafted to exploit a JSON consumer is inert text here — including a __proto__ key, which JSON.parse stores as an ordinary property rather than following it to a prototype.

A malformed line reports as Line 3, column 10 — "o" does not start a JSON value, with the offending line shown beneath it. A valid document reports its depth, node count and any repeated keys instead — the shape information that tells you whether an API response changed.