JSON Diff Tool

Useful when comparing API responses that include timestamps or generated IDs that change every request. In our JSON Diff Tool, click "Ignore keys" and add the key paths you want to skip — for example, createdAt, updatedAt, requestId. Wildcards work too: *.timestamp ignores any key called timestamp at any depth. Programmatically, libraries like deep-diff (Node) and jsondiff (Python) accept a filter function that returns true for keys to skip. Diffs become much cleaner when you mute the noise — you instantly spot the actual changes instead of scrolling past 50 timestamp differences.

YAML and JSON are structurally compatible, so the cleanest path is: convert both YAML files to JSON (using yq on the command line or js-yaml in Node), then diff the JSON. Our JSON Diff Tool has a "Paste YAML" toggle that does this conversion automatically — paste YAML, see the structured diff. Why this works: comments and formatting differences in YAML are noise; you usually care about the data, not whether someone used 2-space vs 4-space indent. For preserving comments, use a YAML-specific tool like yamldiff. For most config-file comparisons, JSON-based diffing is faster.

Our tool colour-codes the output: green for additions (keys in the new file not in the old), red for removals (keys in the old but not the new), and yellow for value changes (same key, different value). You can toggle each colour to focus on one category. The summary panel shows counts: e.g., "+12 added, -3 removed, ~5 changed." For programmatic use, libraries like microdiff return a structured array of {type, path, value} objects you can render or process. Useful in API regression tests where you want to alert only on removed or changed fields.

Browsers can struggle with very large JSON because parsing and rendering both fight for memory. Our tool handles up to about 25 MB per file using streaming JSON parsing and virtualised rendering — you only see the diff lines on screen at any moment, not all 10 million. For files larger than that, do the diff offline with jq or a Python script and just review the output. Tip: pre-sort both files' keys before diffing — different key orders look like changes when they aren't. We also have an "Object key sort" option built in for this exact reason.

Almost always one of three culprits. One: different key order. JSON objects don't have an inherent order, but text-based diffs do — sort both files' keys alphabetically before comparing. Two: trailing whitespace, line endings (CRLF vs LF), or BOM characters at the start of the file. Three: number formatting — 1.0 vs 1, or 1e2 vs 100. Our JSON Diff Tool normalises all of these by default (parsing into objects, then comparing semantically), so it shouldn't false-positive. If you're using a plain text diff like diff on the command line, switch to a structural JSON differ.

Our tool's "Export" button gives you three formats: plain-text patch (similar to a git diff), HTML with the colour-coded view (great for emailing to a teammate), and machine-readable JSON-Patch (RFC 6902) which describes the changes as add/remove/replace operations. Pick whatever suits your use. For team review, HTML is most readable. For CI pipelines that reject regressions, JSON-Patch lets you assert on specific change types. We also support a "share via link" feature that creates a unique URL containing the diff — handy when reviewing changes with non-developer stakeholders.