JSON Formatter

Paste JSON into the left box, drop a .json file on it, or press Upload. The status line under the box checks it as you type: a green line says Valid JSON with the size, the number of keys, the number of array items and the depth, and a red line gives the exact error with its line and column. As soon as the JSON is valid, the right side shows a neatly indented copy.

This json formatter online runs in any modern browser on a computer, tablet or phone. Everything happens on your own device: the text is parsed and formatted by code in the page, and nothing is sent to a server. There is no account, no limit on how often you use it, and files up to 100 MB can be opened.

Beyond plain formatting, the toolbar can minify JSON to a single line, repair almost-JSON copied from code or logs, escape JSON so it fits inside a string, and unescape it again. The output can be read as text, as a folding tree or as a table, and the Path box narrows everything down to the one part you care about.

The page is a json formatter and validator in one. Every change in the input box is checked straight away with your browser's own JSON parser, which follows the same strict rules as JSON.parse in JavaScript. When the text passes, the status line turns green and reads Valid JSON, followed by the size of the text and three counts for the whole document: keys in all objects, items in all arrays, and how deeply the data is nested.

When it fails, a second parser reads the text again to find out where and why. The status line then says Invalid JSON with a plain reason and the position, for example Comma before } (trailing comma) at line 14, column 3, and the box border turns red. Press Show and the cursor jumps to that exact character in the box, so you do not have to count lines yourself.

Line and column both start at 1, and a tab counts as one column. The position is where the parser gave up, which is usually on the problem itself or just after it. A missing comma, for example, is reported at the start of the next value, not at the end of the one before it.

Most errors come from a short list of rules that JSON is strict about. Keys and strings must use double quotes, never single quotes. Every key needs quotes, even a simple one such as id. Items and properties are separated by commas, and there is no comma after the last one.

Only a few kinds of value exist: strings, numbers, true and false, null, and objects or arrays that hold more of them. True with a capital T, None, NaN, undefined and dates written without quotes are not JSON. Numbers are written in plain decimal, with an optional minus sign, fraction and exponent; leading zeros, a leading plus sign and hex are not allowed.

Inside a string, a double quote or a backslash must be escaped with a backslash, and a line break must be written as \n. Comments are not part of JSON at all. The whole text must be one value: one object, one array or one single value, with nothing after it but spaces.

Plenty of text looks like JSON but is not quite: a config file with comments, an object copied from JavaScript code, or a Python dictionary printed in a log. Press Repair, or Try to repair on the red status line, and a forgiving parser reads it and writes proper JSON back into the input box, formatted with your chosen indent.

Repair does not guess at broken structure. A missing closing bracket, a string that never ends, or two separate objects one after the other still fail, and the message tells you where. There is no undo button for the input box, so if you need the original text, copy it somewhere first.

Check repaired data before you rely on it. NaN, Infinity and undefined have no JSON equivalent, so they become null, and comments are dropped because JSON has no place for them. If a value matters, put it back by hand in a form JSON allows, such as a string.

Formatting, also called beautifying or pretty printing, adds a line break after every value and indents each level of nesting so the structure is easy to follow. If you were looking for a json beautifier free online, this is the same job: paste the data, and the indented version appears on the right.

Pick the indent with the three buttons next to Indent. The output changes straight away, and Format writes it out with the new setting. Beautifying never changes the data itself, only the spaces and line breaks between values.

Formatting works by reading the JSON into memory and writing it out again, so a few details come out in a standard form: 1.0 becomes 1, 1e3 becomes 1000, \u00e9 becomes é and \/ becomes /. The meaning is identical; only the spelling changes. If you need the original text byte for byte, keep a copy.

Switch the output to Tree and the JSON becomes an outline you can fold. This json formatter with collapse support opens the first two levels when you start, so you see the top of the data without being buried in it. Click the small arrow next to any object or array to fold or unfold it.

A folded object shows how many keys it holds and a folded array shows how many items, such as 12 items or 5 keys, so you can judge the size of a branch before you open it. Click that badge to open it. Expand all opens every level at once, and Collapse all folds everything below the top.

Keys are shown in purple and work as buttons. Click one and its path, such as items[0].name, goes into the Path box, and the output narrows to that value. Strings are green, numbers blue, true and false amber, and null grey, so types stand out at a glance. A very long array or object shows its first 2,000 children and a note with how many more there are; use the Path box to reach the rest.

Minify (one line) strips every space, tab and line break that sits between values and puts the whole document on a single line. Spaces inside strings are kept, because they are part of the data. Use the page as a json formatter to one line when you need to paste data into an environment variable, a command line, a single log entry, a test fixture or a request body.

Minified JSON is also smaller to store and send. How much smaller depends on the original: the more indentation and line breaks it had, the more it shrinks, while a file that was already compact barely changes. The size in the status line is the size of what is in the input box, so paste the minified result back in to compare.

To go back, press Format and the data is laid out again with your indent. Minify also follows Sort keys A–Z and the Path box, so you can produce one sorted line from just one part of a large file.

Upload accepts files up to 100 MB, and you can drop a file straight onto the input box. Because everything runs in your browser, a large file never has to travel to a server and back, and what the page can handle is set mostly by your device's memory.

A few habits make this json formatter for large data feel quick. Read the status line first: it gives the size, key count, array items and depth before you scroll anything. Then use the Tree view, which opens only two levels and shows at most 2,000 children per branch, instead of the Text view, which has to draw every line. When you find the part you want, type its path or click its key in the tree, and work on that alone.

A desktop browser copes with much bigger files than a phone. If a tab becomes slow, close other tabs, keep the tree folded, and use Minify, Copy and download rather than scrolling through endless formatted lines. Table view shows the first 1,000 rows and up to 50 columns, which keeps it responsive even when the array is much longer.

The Path box shows a single part of the document. Separate keys with dots and put array positions in square brackets, counting from 0. A leading $ or $. is allowed, so simple paths copied from other tools work as they are.

The path applies to everything on the right: the Text, Tree and Table views, Format, Minify, Copy and the .json download. The counts in the status line still describe the whole document. If nothing exists at the path, the output says Nothing at that path.

Paths do not support wildcards, filters or quoted keys in brackets. Keys that contain a dot or a square bracket cannot be reached by path, so read them in the Tree view instead.

Many APIs return a list of records: an array where every item is an object with similar keys. Switch to Table and that array becomes a grid, one row per item and one column per key. Using the page as a json formatter to table this way makes it easy to scan prices, names or IDs down a column.

Columns are collected from all the rows, so an item with an extra key adds a column and an item without it leaves that cell empty. A nested object or array in a cell is shown as compact JSON. Long values are cut off with an ellipsis in the grid but kept in full in the CSV.

Press CSV to download the grid as table.csv, ready for a spreadsheet. Values with commas, quotes or line breaks are quoted properly. The grid and the CSV hold the first 1,000 rows and up to 50 columns. To get an Excel file from your data instead, JSON to Excel does that.

If your array is not at the top, as in the Sample, where it sits under items, type items in the Path box. If the view says it needs an array of objects, the data at that point is a single object, a plain list of values, or a list that mixes objects with other things.

Sometimes JSON has to live inside another string: a field in another JSON document, a string in source code, or a message on a queue. Every double quote then needs a backslash in front of it, and line breaks become \n. A json formatter with escape characters handles that for you in one click.

Escape takes the JSON and outputs it as a single quoted string, with quotes, backslashes, tabs and line breaks escaped. It uses the output if you have just pressed Format or Minify, and otherwise the input exactly as typed. Press Minify first and then Escape for the shortest one-line string.

Unescape works the other way. Paste an escaped string, with or without its outer quotes, and press Unescape: the backslashes are removed and the plain JSON is put into the input box, where it is checked and formatted like anything else. If the text is not a valid escaped string, you see This is not an escaped JSON string.

An escaped string pasted with its outer quotes is valid JSON on its own: the validator sees one string value, and the Tree view shows it as a single green line. That is the sign it still needs Unescape. JSON allows a forward slash to be escaped as \/ but does not require it, so Escape leaves slashes as they are.

Tick Sort keys A–Z and the keys in every object are put in alphabetical order, at every level of nesting. Arrays keep their order, because the position of an item usually means something. The setting affects the output, Format, Minify, the tree and the table.

Sorting helps when you compare two versions of a file. Two JSON documents with the same data but keys in a different order look completely different in a text diff; sorted, the real changes stand out. The order is by character code, so capital letters come before small ones and digits before letters: Zip sorts before city.

This is a json formatter free to use as often as you like, with no sign-up, no account and no daily limit. Every button you see is open to everyone.

Your JSON is parsed, checked, repaired and formatted by code running in this page, on your own device. It is not uploaded, and files you open are read by the browser, not sent to us. That matters for API responses with tokens, customer exports and private config files. Only your settings are kept, in this browser's local storage; the JSON itself is not saved anywhere.

Reading a big document is hard enough without things moving around it. There are no pop-ups or ads inside the formatter: the toolbar, the input box and the output panel are all tool, and nothing is placed between your data and the result. If you wanted a json formatter without ads getting in the way of your work, the tool area is kept clean for exactly that reason.

There is also nothing to install: no extension, no desktop app and no plugin. Bookmark the page and it opens with your last indent, view and sort settings.

Need the data in a spreadsheet? JSON to Excel turns an array of records into an Excel file, and CSV to PDF makes a printable page from the CSV you downloaded in Table view. Writing documentation for your API? Word Counter checks its length and Case Converter tidies the headings. Need filler text for a mock response? Lorem Ipsum Generator makes some in a click.