About CompareFiles

CompareFiles compares two JSON documents — API responses, config files, exported payloads — and shows exactly what changed: fields added, fields removed, values modified, and types changed. It exists because most JSON diff tools either upload your data to a server or bury the actual differences under noisy, unreadable output.

Privacy is the architecture, not a policy

Parsing and comparison run locally in your browser, off the main UI thread in a Web Worker, so large documents stay responsive without ever leaving your machine. There are no accounts, no analytics, and no telemetry. Payload contents, paths, values, notes, headers, and report contents are never written to logs or error monitoring.

The one exception is opt-in: fetching a response from a URL or cURL command goes through a secure server-side proxy so browser CORS restrictions don't block the request. That proxy strips credentials before returning a response, enforces HTTPS and hostname allowlisting in production, and never stores what it fetches.

Built for real comparison work, not toy diffs

Object key order never affects results, duplicate array items are compared as multisets in unordered mode, and every field is tracked by its exact RFC 6901 JSON Pointer path — so a rename three levels deep in a nested array is reported precisely, not approximated. A separate Structure Schema comparison checks shape consistency independent of the values themselves, which is useful when you care that two API responses agree on their contract even if the actual data differs.

Reports are Markdown, reviewed in an in-page preview before you download or copy them, and carry a standing warning that the values inside may be sensitive — because the tool has no way to know what you pasted into it.

See the documentation for the full feature set, including array modes, ignore-path patterns, and secure remote fetch.