JSON Transform Tools

18 tools · free · nothing uploaded

JSON in, JSON out. These are the operations you would otherwise open a Node REPL for — and the two that matter most for version control are sorting keys and flattening, because both turn an unreviewable diff into a readable one.

Start here

Flatten Nested objects to user.address.city keys, in dot or bracket notation. Sort keys Recursive alphabetical order, which is what makes two API dumps diffable. Deep merge Merge documents with an explicit rule for what happens when both sides set a key. Convert key case camelCase to snake_case and back across a whole document, collisions flagged. Filter array Keep or drop records with a readable condition, when jq syntax is more than you need. Group & aggregate GROUP BY with count, sum, average, min, max and median, over an array of objects.

Everything else in Transform

Unflatten — Dot / bracket keys back to nested objects.
Extract — Pull keys or values by literal, glob or regex.
Escape — Backslash-escape into a JSON string literal.
Unescape — Turn an escaped string back into JSON.
Rename keys — Bulk-rename or drop keys by exact name or regex.
Remove nulls — Strip nulls, empty strings, arrays and objects.
Dedupe array — Remove duplicate elements, whole-value or by key.
Array ↔ object — Key an array by a field, or split an object back out.
Sort array — Order elements by one or more fields, stable and type-aware.
Set operations — Union, intersection and difference of two arrays.
Redact — Mask or remove secrets by key name, pattern or value.
Prune — Cap depth, array length, string size and keys per object.

Why these run in your browser

There is no server-side processing here and no upload step: the page loads, and from then on the parsing, converting and rendering happen in this tab using your machine's CPU. That has three consequences worth knowing about. Your data stays on your machine, which matters when the JSON is a production payload with real customer records in it. There is no file-size limit imposed by us, only by how much memory your browser will give the tab. And once the page has loaded, it keeps working with the network off.

The trade is that very large documents are bounded by browser memory rather than by a beefy server — somewhere past 50 MB a textarea gets sluggish, and dropping the file in is better than pasting it. The privacy page spells out what is and is not collected.