How the comparison works
Both valid inputs are parsed and re-serialized with the same indentation. The result is compared line by line using a longest-common-subsequence diff. If one side is invalid, the page falls back to raw-text comparison and reports the parse status. Added lines are highlighted green, removed lines red, and unchanged context stays neutral.
This is a textual diff of normalized JSON when both sides parse, not a semantic JSON Patch. Key order differences still show as changes, even though the objects are semantically equal. If key order is noise, sort keys on both sides before comparing.
- Useful for reviewing API response changes between environments.
- Catches missing fields and type changes (
1vs"1"). - Does not ignore key order or numeric formatting differences automatically.
- For semantic diffs, use a structural JSON diff library that reports paths.
Reading the result
A line that appears on both sides with different values is shown as a removal followed by an addition, which is how line-based diffs represent modifications. Look for pairs of red and green lines at the same position to identify changed values.
If the two documents are large and the diff is mostly noise, compare a sorted, minified version first to confirm whether any values changed at all.