Reverse Each Word Without Reordering the Sentence
Reverse letters inside word-like runs while preserving punctuation and separators in place.
Reverse Each Word changes the letters and numbers inside each word-like run while preserving where those runs occur in the line. It is useful for word puzzles and character-order experiments. It is a different operation from reversing the sequence of words, and the difference becomes obvious as soon as punctuation or multiple spaces appear.
In this guide
What counts as a run
The implementation finds consecutive Unicode letters, combining marks, and numbers, then reverses each matched run by grapheme clusters. Everything outside those runs stays at its original position. “red blue” therefore becomes “der eulb,” while the space remains between the two runs. “red, blue!” becomes “der, eulb!” because comma, space, and exclamation mark are separators rather than part of either matched run.
Apostrophes and hyphens also act as separators. A term such as “well-known” is reversed in two parts, not as one long string. A contraction is split around its apostrophe. Test such text if the puzzle requires a specific rule about what counts as one word.
Inputred, blue!Output
der, eulb!
Compare neighboring tools
The separate Reverse Words tool moves whole nonspace runs to different positions; it leaves the letters in each run readable. The homepage Reverse Text tab reverses UTF-16 units within each line, so punctuation and spaces move too, but supplementary characters and combining marks can break. Reverse Each Word preserves the sentence skeleton and changes only the interior of letter and number runs using the newer grapheme helper.
Use a sample with punctuation and two spaces to select the correct operation. The apparent equivalence of these tools on a single word disappears on a real sentence. A multi-line input remains multi-line here because the replacement takes place within the original string. If you use the homepage tabs instead of dedicated tool pages, remember that their older implementation can apply different spacing and Unicode rules.
Unicode details
Grapheme-aware reversal helps a letter and its combining accent move as one visible unit in browsers that support Intl.Segmenter. The word boundary rule itself is a simple regular expression, however. It is not a full linguistic word segmenter, and scripts without spaces or with complex joining behavior may need manual review.
The tool does not normalize characters before reversing. Visually similar inputs can have different underlying code-point sequences and therefore different matching behavior. For editorial text, inspect a short sample from the actual source rather than assuming every script follows English word boundaries.
Practical use
Paste the original, check one representative line, and copy the result only after examining separators. Keep the original alongside the result if you will need to restore the phrase. The operation is reversible for many simple words, but unusual combining sequences and later edits can make recovery less straightforward.
For a worksheet, decide whether numerals should reverse too. The tool includes number runs, so “item 123” becomes “meti 321.” If numbers are identifiers, that may be undesirable. It also leaves an underscore between runs, so a code-like token may become two independently reversed parts. These examples make the operation better suited to deliberate word play than to transforming structured data. Review sample identifiers before processing a whole file.
- Keep punctuation where it is by choosing this tool.
- Use Reverse Words if word positions must change.
- Review language-specific text after conversion.
Frequently asked questions
Will spaces move?
No. Spaces are outside the matched letter and number runs and remain in place.
Are hyphenated words reversed as one unit?
No. Each side of the hyphen is matched and reversed separately.
Does the tool reverse emoji?
No. Emoji generally fall outside its letter, mark, and number runs.