Compare Text Revisions Line by Line
Compare two short revisions by line while distinguishing textual differences from changes in meaning.
Open the tool ↓Keep original and revision in the correct fields
The left field is the original text and the right field is the revised text. Reversing them swaps the interpretation of added and removed lines. Use plain text copied from the same representation rather than comparing one rendered page with another page’s source markup.
Text Diff Checker
Open the dedicated comparison view for its line-by-line output; large comparisons can consume substantial browser memory.
Open Text Diff Checker →Follow a two-line example
Put these lines in the original:
name: Ada
status: draftPut these lines in the revision:
name: Ada
status: readyThe unchanged name line is retained. The changed status line is represented as one removed line and one added line. That does not mean two independent editing actions occurred; a line-level diff represents replacement through removal and addition.
Read the comparison at the right level
The implementation uses a longest-common-subsequence comparison of exact lines. It does not perform character-level highlighting, parse JSON meaning, or decide whether two documents express the same idea. Moving a block can appear as a removal and an addition rather than a special “moved” operation.
- Paste a small, relevant excerpt into each side.
- Read unchanged lines for context around each difference.
- Verify which side contains the intended final wording.
- Repeat on another excerpt if the full documents are very long.
Do not ignore whitespace accidentally
Spaces, tabs and line breaks are part of the comparison. A trailing blank line can produce an extra changed line even if the documents look alike in an editor. When comparing structured data, format both with a consistent convention first if whitespace differences distract from the task.
JSON Formatter can make a JSON comparison easier to read, but parsing may normalize numeric representation or duplicate keys. Keep the originals if those details matter. Formatting before diffing is an analysis choice, not a lossless transformation guarantee.
Understand performance and confidentiality
The line-matrix algorithm uses memory proportional to the product of the two line counts. Very large comparisons can slow or freeze the tab; the dedicated view does not make the computation unlimited. Start with a short excerpt instead of pasting enormous logs.
Comparison runs in the browser. That does not remove clipboard, screen-sharing or device-level risks for sensitive text. Use fictional or redacted samples when learning the workflow and review the intended final revision before copying it into another system.