UtilityPilot.Find a tool

Text & Lists

Text Diff — Compare Two Texts

Compare text from two versions of a draft, message or configuration. Choose line or word comparison to see added, removed and unchanged passages, then copy or download the comparison report.

Processed in your browserPrivacy details ↗

256 KiB per version

Adjust your options

No sign-up. No upload.

What does a text comparison tool do?

Text comparison identifies changes between two sequences of text. An earlier draft and an edited draft may look similar at a glance, but a diff makes insertions and deletions explicit. It is useful when you need to review an editor’s changes, check a rewritten paragraph or find which line changed in a small configuration.

UtilityPilot compares the text you paste, including its order. It does not interpret meaning, score similarity or establish whether one version is better. A change from “may” to “must” is reported as a text edit; deciding its significance remains part of your review.

How to compare two texts

  1. Paste the original or earlier text into Input A.
  2. Paste the revised text into Input B. The direction matters: additions belong to B and removals come from A.
  3. Select Lines for structural changes or Words for smaller wording edits.
  4. Leave Normalize spaces and line endings off unless those differences should be ignored.
  5. Select Compare text. Read the labelled result, then copy or download the text report if needed.

The source inputs remain unchanged. If you change either input or an option, run the comparison again. An empty A can show the contents of B as additions; an empty B can show A as removals. Two empty inputs have no differences, which does not tell you whether the intended text was pasted.

How the text diff finds differences

The processor uses the bundled jsdiff sequence-comparison library. Lines mode compares line tokens. Words mode uses word-and-space tokens, keeping whitespace in the comparison rather than discarding it. This lets the result show textual edits without treating an entire paragraph as one indivisible item.

The algorithm produces a sequence of unchanged, removed and added passages. A replacement is represented by a removal and an addition, not a separate semantic “changed idea” operation. Repeated words or repeated lines can allow several valid ways of aligning the versions, so the particular highlighted path is not a record of the author’s keystrokes.

The tool has no character-by-character selector, ignore-case switch, automatic merge or plagiarism analysis. Capitalization is significant. To understand whether a punctuation or capitalization edit is intentional, read the original and revised passages around the highlight.

Reading additions, removals and changed text

  • Added: text found in the revised sequence B that the comparison inserts.
  • Removed: text from A that the comparison deletes.
  • Unchanged: aligned text shared by the compared sequences.

The result uses written labels as well as color. Copy result and Download produce a plain-text report with Added, Removed and Unchanged labels. This is a review report, not a patch file intended to be applied by Git or another patch utility.

Input

Input A: The draft is ready.
Input B: The final draft is ready.

Result

In Words mode, “final ” is added.
In Lines mode, the old line is removed and the revised line is added.

Words mode is useful for this small edit because it keeps the shared sentence visible around the inserted word. Lines mode is often clearer when several lines have been inserted or removed together.

When to normalize whitespace

Enable Normalize spaces and line endings when repeated spaces or tabs are accidental noise. The comparison first converts CRLF and CR line endings to LF, collapses runs of ordinary spaces and tabs to one space, and removes the resulting ordinary space at each line edge.

It does not remove blank lines, collapse all Unicode whitespace or make the two raw inputs byte-identical. For example, “A” followed by two spaces and “B” can compare equal to “A B” with normalization enabled. A non-breaking space is not handled by that same ordinary-space rule.

Keep normalization off when indentation, tabs or exact spacing matter, especially in configuration or structured text. If you intentionally clean a document first, use Text Cleaner on a copy, then compare the versions under matching rules.

Practical text-comparison workflows

Review an edited draft

Place the submitted draft in A and the edited copy in B. Start with Words, read each removal and addition in context, and check names, dates and claims against the source material. A diff helps locate edits; it does not fact-check them.

Check configuration or pasted data

Use Lines with whitespace normalization disabled. Inspect changed keys, values and blank lines. This is a plain-text comparison, not a parser: equivalent JSON objects can look different because of member order or formatting. To make JSON indentation consistent first, use JSON Formatter, while keeping an original copy.

Verify a repeated replacement

After using Find and Replace Text, compare the original with the output to check whether a short search also altered longer words. Whole-word matching can prevent some unintended replacements, but reviewing context is still useful.

Text diff vs comparing lists

Text Diff answers “What changed in this sequence?” Compare Lists answers “Which unique entries are in both lists or only one?” These are different questions.

If A contains apple then pear, and B contains pear then apple, their membership is the same but their sequence differs. Choose Text Diff when order or repeated occurrences matter. Choose Compare Lists when finding shared or missing entries is the goal, and consult the list comparison and text difference guide for a worked example.

Check text length after comparing revisions

A shorter-looking diff does not establish a shorter draft: replacement words can be longer, and blank lines can change the structure without changing word count. After reviewing additions and removals, paste the final draft into Word Counter to check a writing limit or its basic text-analysis summary.

For a constrained field, inspect the character count. For a paragraph-based assignment, inspect paragraph blocks. If copied spacing is the problem, use the plain-text formatter and compare the cleaned version with normalization off so those spacing edits remain visible.

Comparison limits and common mistakes

Each side accepts up to 256 KiB of UTF-8 text. The diff calculation is bounded to 1.5 seconds and a maximum edit length of 10,000. Very different inputs can reach the comparison limit even when their byte size is acceptable; compare corresponding smaller sections if that happens.

Paste text from PDF or Word documents yourself; this page does not extract document text, compare page layout, inspect images or preserve rich formatting. Keep A and B in a consistent earlier/later order. If the result says No differences found with normalization on, it refers to normalized text, not an exact raw-file identity check.

Processing and privacy

Text Diff compares the two versions in page memory; neither version is uploaded or retained after you leave. Input stays in browser memory until you reset or leave the page. Copy and download run only when you choose them. No tool input is sent to a processing service. Website pages and scripts still make ordinary hosting requests; the privacy policy separates those from tool processing.

Frequently asked questions

Does this compare the meaning of two texts?

No. It compares text sequences by lines or word-and-space tokens. It does not provide semantic similarity, authorship or plagiarism judgments.

Why does changing one word mark a whole line?

Lines mode uses whole lines as comparison units. Select Words and compare again to inspect smaller edits within that line.

Can I ignore capitalization?

There is no ignore-case option in Text Diff. Letter case is part of the comparison; whitespace normalization does not change it.

How this tool was checked

Text Diff was checked with line and word modes, whitespace normalization, repeated lines that allow several valid alignments, insertions at the start and end, and 256 KiB inputs. See the testing method or report a failure.