How does the LCS algorithm help?

Longest common subsequence matching finds the largest set of tokens that appear in the same order in both texts. That lets the diff align moved or reordered content correctly, rather than marking everything after the first change as different.

What does Ignore whitespace actually ignore?

In word and character modes it removes all whitespace before comparison; in line mode it collapses runs of whitespace within lines to single spaces. Together with the comparison, the original text is still displayed unchanged — only the matching is relaxed.

Can I use the unified output with git apply?

The unified diff uses --- a / +++ b headers and standard @@ hunks, but it is intended for reading and archival. Applying it to real files requires matching file paths and exact context, so prefer git diff output for actual patching.

Why does an identical comparison show only headers?

When every token matches under the active options, there are no changes to report, so the unified output contains just the --- a and +++ b lines. The green Texts are identical badge appears above it to make the result explicit.

How are added and removed counts computed?

Each token-level operation produced by the LCS backtracking is tallied: added and removed tokens are counted separately, and unchanged tokens count matches. The three stats update live as either text is edited.

Developer

Text Diff

Compare text at line, word, or character level with unified diff output.

diffcomparelinewordcharacter

Loading tool…

Recommended tools

Cloudflare

Free CDN, DNS, and edge hosting for developers.

Namecheap

Affordable domains and SSL certificates.

Affiliate links — we may earn a commission if you sign up.

About the Text Diff

The Text Diff tool compares two blocks of text at three granularities — line, word, and character — using a real LCS (longest common subsequence) algorithm rather than a naive positional compare, so reordered lines still align correctly and only genuine changes are flagged. Text A and Text B sit side by side with live line counts, and a tab strip switches between Line diff, Word diff, and Character diff views. The visual pane color-codes everything: added content in green, removed content in red with strikethrough in word and character modes, unchanged text in gray, with per-line + / - prefixes in line mode. Two switches change the comparison semantics — Ignore whitespace collapses all spacing before comparing, and Ignore case lowercases both inputs — and when the texts turn out identical under the current options, an explicit Texts are identical badge appears instead of an empty pane. A Unified diff panel renders standard patch format with @@ hunk headers and three lines of context, downloadable as diff.patch. It is ideal for reviewing prose revisions, comparing config snapshots, or auditing exactly what changed between two versions.

Hand-written guide

Examples

Input

Text A: The quick brown fox
jumps over the lazy dog.
A classic pangram for testing.
Line four is unchanged.
Line five will be removed.

Text B: The quick red fox
jumps over the sleepy dog.
A classic pangram for testing.
A brand new line was added here.
Line four is unchanged.

Output

--- a
+++ b
@@ -1,5 +1,5 @@
-The quick brown fox
-jumps over the lazy dog.
+The quick red fox
+jumps over the sleepy dog.
 A classic pangram for testing.
+A brand new line was added here.
 Line four is unchanged.
-Line five will be removed.

Note: The built-in samples in Line diff mode — the LCS alignment matches the reordered Line four correctly, and the unified view shows one hunk.

Input

Text A: cat
Text B: cut (Character diff mode)

Output

--- a
+++ b
 c
-a
+u
 t

Note: Character mode isolates the single edited letter: a struck through in red, u highlighted in green, with c and t as unchanged context.

Input

Text A: SELECT id FROM users
Text B: select id from users
(Ignore case on)

Output

--- a
+++ b

Note: With Ignore case enabled the texts match, the unified diff contains only headers, and the Texts are identical (case-insensitive) badge appears.

How to use

1
Paste the original into Text A (original) and the changed version into Text B (modified).
2
Pick a granularity with the tabs: Line diff, Word diff, or Character diff.
3
Toggle Ignore whitespace and Ignore case to tune what counts as a difference.
4
Read the visual pane — green added, red removed, gray unchanged.
5
Copy or download the Unified diff (diff.patch) for use with patch tooling.

Common use cases

Reviewing prose edits sentence by sentence without missing a moved paragraph.
Comparing two configuration snapshots to audit exactly which keys changed.
Spotting a one-character typo inside a long token with Character diff mode.
Checking that a formatter only changed whitespace by enabling Ignore whitespace.
Reviewing translations where word-level changes matter more than whole lines.
Generating a patch-style report of changes for a release note.

Best practices

Choose line mode for documents, word mode for sentence-level edits, and character mode for typos inside words.
Enable Ignore whitespace before comparing code that has been auto-formatted — indentation noise hides real edits.
Use Ignore case when comparing case-insensitive formats like SQL keywords or config keys.
Trust the stats (Added / Removed / Unchanged) for a quick summary before reading the visual diff.
Remember word and character modes emit a single-stream unified view, while line mode produces @@ hunks with context.
Load the sample first to learn the color coding, then paste your own texts.

Tips

Use Character diff mode when a test failure message differs by one byte — it finds the needle instantly.
Turn on Ignore whitespace before comparing JSON or code that different tools have pretty-printed.
Watch for the Texts are identical badge instead of scanning empty panes after relaxing both switches.
Download diff.patch for records — it includes hunk context that a screenshot cannot capture.

Frequently asked questions

Line mode splits both texts on newlines and compares whole lines. Word mode tokenizes into words and whitespace, so edits inside a line are isolated. Character mode compares individual characters, which is the finest granularity and ideal for spotting typos or casing changes.

Explore more developer tools

Browse the full collection of developer tools on the hub, or jump back to all categories.

All Developer tools All categories