Keep an accented character intact
The decomposed text e followed by a combining acute accent forms one grapheme cluster. Character reversal keeps the accent attached instead of moving it to the neighboring letter.
Reverse grapheme clusters across the entire text, including line breaks.
Maximum 100,000 UTF-16 code units; line modes allow 10,000 lines. Character modes require Intl.Segmenter. Grapheme clusters keep combining marks and supported emoji sequences together. This does not mirror glyphs or change right-to-left rendering rules. Processing stays in this browser; text is not submitted to a server.
Text is processed locally in your browser. No tool input is sent to a remote processing service.
How to use the Text Reverser
- 01
Paste the text to reverse.
- 02
Choose all character clusters, characters within each line, word order, or line order.
- 03
Inspect the result. Use result as input when you want to apply another transformation.
- 04
Copy or download the reversed plain text. Reset restores the default mode and clears the text.
Examples
Reverse words without rearranging whitespace slots
“one two\tthree” becomes “three two\tone”. The double-space slot and tab slot stay where they were while non-whitespace tokens change order.
Reverse the itinerary, not each destination
For Paris, Rome, and Oslo on separate lines, line mode yields Oslo, Rome, Paris. The letters within the city names remain unchanged.
Choose the unit before reversing
All-character mode reverses the sequence of grapheme clusters across the entire input. Line breaks are part of that sequence, so line order also reverses. Each-line mode keeps the lines in place and reverses only their contents. Word mode treats every non-whitespace run as a token, including attached punctuation, and reverses those tokens through the existing whitespace slots. Line mode treats each complete line as an indivisible record.
Grapheme clusters are practical visible-character units
A JavaScript string can represent one visible symbol with multiple code units or code points. Character modes use the browser’s Intl.Segmenter grapheme segmentation, which keeps combining marks and supported joined emoji sequences together. Segmentation follows the Unicode data bundled with the browser, so newly introduced sequences may differ on older platforms. A browser without Intl.Segmenter receives a visible error rather than a silent code-point fallback.
Reversal does not mirror writing
The tool changes logical text order. It does not flip glyph shapes, turn text upside down, reverse the pixels in an image, or override bidirectional rendering. Arabic and Hebrew can look surprising after logical reversal because the browser continues to apply its normal direction and shaping rules. Punctuation stays attached to its word in word mode; the tool does not reinterpret sentence grammar or move a period to produce a grammatical reversed sentence.
Line endings and reversibility
Line modes normalize Windows CRLF and old-style CR to LF. A trailing newline becomes a final empty line, and reversing line order moves that empty line to the beginning. Most ordinary inputs return to their starting order after the same reversal is applied twice, but newline normalization is not undone. Unicode grapheme boundaries can also depend on adjacent text, so unusual boundary-sensitive sequences are not promised to round-trip exactly. Keep the original text if byte-for-byte preservation matters.
What to know
- Maximum 100,000 UTF-16 code units; line modes accept at most 10,000 lines.
- Character modes require Intl.Segmenter. Logical reversal does not mirror glyphs or bypass bidirectional display.
Common questions
Does it preserve emoji skin tones and family emoji?
Character modes keep a sequence together when the browser’s grapheme segmenter recognizes it as one cluster. Support depends on the Unicode data in the browser.
Can I reverse only the order of lines?
Yes. Select Line order. Each line’s contents remain intact, while the first line becomes the last.
Is reversed text a secure encoding?
No. Reversal is easy to undo and should not be used to protect private information.