JSON formatter and syntax validator

Format a response so you can scan its structure

A one-line API response is hard to scan when you need to check which fields belong inside an object or array. Paste the JSON, choose 2 spaces, 4 spaces or Tab, and select Validate and format. The formatter adds line breaks and indentation outside strings so the nesting is easier to follow.

For example, {"name":"Mina","active":true} becomes a multi-line object with name and active on separate lines. Choosing Compact to one line removes formatting whitespace between tokens. A space inside a value such as "Mina Lee" stays there because it is part of the string, not indentation.

The input can be an object or array, or a single string, number, boolean or null value. Formatting changes how the JSON is laid out. It does not rename fields, change values or decide whether the data is useful to the receiving service.

Use the line and column hint to find syntax errors

A validator can point to where parsing stopped, but it cannot know what you intended to write. The reported location may be just after the original mistake. Start there, then check the preceding value, comma, colon and quotation marks.

In this example, a comma is missing between the status and data fields. The error points to the opening quote of data at line 1, column 16. Add the comma yourself, then run the formatter again. The corrected JSON and its compact result are shown below.

Object keys and strings need double quotes. JSON also rejects comments, single-quoted strings, trailing commas, NaN and Infinity. A leading zero in a number is invalid, so use a string such as "001" only if the receiving API expects that identifier to be a string.

Valid JSON does not guarantee an API accepts it

Syntax validation answers whether the text follows JSON grammar. It does not check required fields, field names, value types, date formats or application rules. For example, {"active":"true"} and {"active":true} are both valid JSON, but the first value is a string and the second is a boolean. If an endpoint expects a boolean, formatting cannot change a quoted string into one.

Duplicate object keys need extra care. This formatter preserves every occurrence in its original order and does not choose a value. RFC 8259 recommends unique names; it also notes that receiving software can behave unpredictably when names repeat. Check the source data and the API contract instead of treating a successful format as a decision about which value is correct.

Number tokens and string escapes stay as written

The formatter works from JSON tokens rather than converting the document through JavaScript numbers. It preserves the spelling of numbers, key order, duplicate keys and escapes inside strings. For example, the number 9007199254740993 remains those same digits in the output, and 1e3 is not rewritten as 1000.

That is a property of this formatter's output, not a promise about the next program that reads it. A receiving system may use a numeric type with a smaller exact range or different precision. RFC 8259 describes the interoperability limits of commonly used double-precision numbers. If a value is an identifier, use a JSON string only when the receiving contract defines it as a string; do not quote every number as a workaround.

Whitespace outside strings is formatting and can be removed in compact mode. Characters inside strings are part of the value, so the formatter keeps spaces and escape spellings such as "\u0061\n" rather than decoding and rewriting them.

Work within the input and output limits

One input can contain up to 100,000 UTF-16 code units, 64 nested object or array levels and 5,000 values. Objects and arrays each count as a value. A supplementary character such as many emoji uses two UTF-16 code units, and a combined emoji can use more. The formatted output can be up to 2,000,000 UTF-16 code units. If formatting exceeds that limit, reduce the indentation or compact the JSON; compact output is bounded by the smaller input limit.

Error lines and columns start at 1. A tab counts as one column, and the column calculation counts UTF-16 code units. A supplementary character may therefore count as two columns even when it looks like one symbol. Use the location to narrow down the area, then inspect the text around it.

Your JSON stays in the browser and is not uploaded or saved automatically. Reloading clears the page. Editing the input or changing indentation clears the previous result, so run it again before copying or downloading. A download creates a separate JSON file and does not overwrite your source file.

When comparing two versions, first format both with the same indentation. The text comparison tool can show line changes, but it does not decide whether two JSON documents mean the same thing. Use the word counter when you need a text length rather than a syntax check.

Questions about JSON formatting

References

• RFC 8259: The JSON Data Interchange FormatThe IETF specification for JSON syntax, values, objects and numbers; Sections 4 and 6 describe duplicate-name and numeric interoperability considerations.