JSON Diff

A JSON diff compares two JSON documents and highlights exactly what changed between them — which keys were added, removed, or modified, and which array elements were inserted or deleted. Our tool parses both documents strictly against the JSON specification (RFC 8259), pretty-prints them with a consistent indentation, and then runs a Myers shortest-edit-script algorithm on the resulting lines. The output shows added lines in green, removed lines in red, and unchanged lines in neutral — with per-side line numbers so you can locate every change instantly.

Yes — 100% private. Both documents are compared entirely in your browser using JavaScript. Nothing is ever uploaded to a server, so it is safe to compare API responses, configuration files, database dumps, or auth payloads. The tool also works offline after the page loads, and your inputs are only stored in your own browser's localStorage so you can pick up where you left off.

Green (with a + marker) shows lines that exist in JSON B but not in JSON A — content that was added. Red (with a − marker) shows lines that exist in JSON A but not in JSON B — content that was removed. Neutral lines are unchanged and appear in both documents. A single line changing value (for example "version": "1.0.0" becoming "1.1.0") appears as one red removed line followed by one green added line.

Yes. Because the tool pretty-prints both documents with the same indentation before diffing, changes deep inside nested objects and arrays show up as localized line changes — only the affected lines are marked. Array insertions and deletions are detected correctly (for example, removing an element from a list shifts the following lines rather than marking everything as changed). Like git diff, the tool preserves each document's key order: if you reorder keys between the two documents, those lines can appear as changes — sort keys consistently on both sides (our JSON formatter has a Sort keys option) before comparing for the cleanest result.

The most common reasons are: (1) whitespace differences — our tool normalizes indentation, so that is rarely the cause; (2) object key order — the tool preserves each document's key order, so reordered keys can appear as changes; (3) number formatting — 1.0 and 1 are different JSON values even though they look similar; (4) type changes — "5" (a string) and 5 (a number) are different. Enable consistent formatting in both documents before comparing for the cleanest diff.

Paste both documents into the two panels, or use the Import button on each side to load .json files directly from disk. The diff engine handles large documents efficiently — it keeps the common prefix and suffix intact and only highlights the differing middle section, even for very large files. You can also copy the unified diff text or download it as a .diff file for sharing with your team.

A plain text diff (like diff on the command line) compares files character-by-character or line-by-line, so formatting differences and key reordering can produce noisy results. A JSON diff first parses and normalizes both documents, so it understands structure — identical values are matched even when formatted differently, and changes are reported in terms of JSON structure (added/removed/modified keys, array elements). This makes JSON diffs much more readable when reviewing API payloads or config changes.

Yes — completely free, with no registration, no sign-up, no premium tier, and no usage limits. All features are available: two-panel comparison, color-coded unified diff, add/remove/unchanged statistics, identical-document detection, copy-diff, download as .diff, swap sides, per-side file import, and clipboard paste.