How to Fix Invalid JSON: Common Errors and Solutions

Trailing commas

JSON does not allow a comma after the last property in an object or the last item in an array. This is the single most common error in hand-edited configs and in JSON generated by language models, because many programming languages (and JSON5) do permit trailing commas. Remove the comma that sits immediately before a closing } or ].

{ "a": 1, "b": 2, }   <- invalid
{ "a": 1, "b": 2 }    <- valid

[ 1, 2, 3, ]          <- invalid
[ 1, 2, 3 ]           <- valid

Single quotes instead of double quotes

JavaScript and Python both allow single-quoted strings, but JSON requires double quotes for every string and every key. If you copied an object literal out of code, this is usually the first thing to fix. Replace each pair of single quotes with double quotes.

{ 'name': 'Ada' }     <- invalid
{ "name": "Ada" }     <- valid

Unquoted or missing keys

Every object key in JSON must be a double-quoted string. Bare keys like {name: "Ada"} are valid JavaScript but invalid JSON. This often appears together with single-quote errors when JSON is pasted from source code. Wrap every key in double quotes.

{ name: "Ada" }       <- invalid
{ "name": "Ada" }     <- valid

Bad escapes and control characters

Strings can break JSON in subtler ways. A literal newline or tab inside a string must be escaped as \n or \t. A backslash must be doubled as \\, and a double quote inside a string must be escaped as \". Unescaped control characters and stray backslashes produce 'Bad control character' or 'Bad escaped character' errors. This is common when JSON is built by string concatenation instead of a proper serializer.

{ "path": "C:\Users\Ada" }     <- invalid (lone backslashes)
{ "path": "C:\\Users\\Ada" }   <- valid

{ "note": "line1
line2" }      <- invalid (raw newline)
{ "note": "line1\nline2" }      <- valid

Wrong literals, comments, and trailing junk

Several smaller mistakes cause parse failures. JSON literals must be lowercase, so True, False, and None (Python-style) are invalid — use true, false, and null. Comments are not allowed in standard JSON, so // and /* */ must be removed. Numbers cannot have leading zeros, a leading plus sign, or values like NaN and Infinity. Finally, a document must contain exactly one top-level value — extra characters after the closing brace, such as a stray semicolon or a second object, will fail.

{ "ok": True, "x": NaN }   // <- invalid
{ "ok": true, "x": 0 }      <- valid (no comment, valid number)

A reliable repair workflow

When you face a broken document, do not guess. Paste it into a JSON Validator to get the exact line and column of the first error, then fix errors one at a time from the top — later errors frequently disappear once the first is resolved, because the parser was misreading everything that followed. A JSON Editor with inline highlighting makes this even faster by flagging problems as you type.

For payloads that are mangled in predictable ways — trailing commas, single quotes, markdown fences from an LLM — a JSON Repair tool can apply the common fixes automatically, after which you re-validate. Everything runs in your browser on JsonDrift, so you can safely repair JSON that contains secrets or customer data without it leaving your machine.