Reverse Text Tool

It turns text round, and it lets you say exactly what "round" means. Seven modes cover the reversals people actually want, from flipping every character to putting the last line first, and a separate setting decides what counts as a character in the first place. Everything runs on your device, not our servers, so a log file, a customer list or a draft you would not send to a website is safe here.

The seven modes

Each row shows what happens to the same short input, so the difference between the two that people mix up most - reversing the letters inside words, and reversing the order of the words - is visible rather than described.

Reversing twice gets you back

Every mode except the delimiter one is its own undo: reverse the result again and you have the original, character for character, including the spacing and the line endings. That is not a convenience, it is the strongest correctness check a tool like this has, and it is the check most implementations fail the moment there is an emoji or a Windows line ending in the text. The Reverse again button beside the result sends the output straight back into the input so you can try it.

A string in a browser is stored as a list of UTF-16 code units, and the one-line reversal every tutorial teaches - split, reverse, join - reverses that list. For English that is fine. For everything else it produces output that looks like a bug report.

What actually goes wrong

This tool reverses grapheme clusters, which is the Unicode standard's own definition of what a reader would call a character. All six cases above come out unchanged, and the tool's own test suite checks that against the browser's built-in segmenter across hundreds of thousands of randomly generated strings - including deliberately malformed ones.

The setting under the mode picker chooses which of the three definitions the reversal uses, and the counts under both text boxes show all three at once so the difference is never a surprise. A thumbs-up with a skin tone is one character, two code points and four code units.

• Characters (grapheme clusters) — the default, and the right answer almost always. It is what a person means by "character", what a text-message limit really counts, and the only one of the three that keeps emoji, flags and accents intact.
• Code points — one entry per Unicode scalar value. This is what Array.from and the spread operator give you in JavaScript, and it is what the better online reversers use. It keeps astral characters in one piece and still breaks flags, skin tones and combining marks.
• Code units — one entry per UTF-16 unit - what String.length counts and what split("") produces. It is offered deliberately, because "what does my buggy reversal actually output" is a real question, and this is the only place you can see the answer next to the correct one.

Switching between them re-runs the reversal immediately, so the three answers can be compared on your own text in a few seconds. That comparison is the reason the option exists.

The four modes that reverse an ORDER rather than characters all follow one rule: the content moves and the separators do not. That single rule is what keeps indentation, blank lines and line endings exactly where they were.

• Line order — recognises Windows CRLF, classic Mac CR, Unix LF and the Unicode line and paragraph separators. Each line comes back with the ending it arrived with, a file that ended in a newline still ends in one, and a blank line in the middle stays a blank line rather than migrating to the top.
• Word order — runs of whitespace stay in their slots, so a double space between two words stays a double space in the same position, and leading indentation is still leading indentation. Tools that move the spacing along with the words turn an aligned column into a ragged one.
• Sentence order — uses real sentence detection rather than splitting on full stops, so "Dr. Smith went to the U.S.A. in 2020." is two sentences and not five. Decimals, initials and abbreviations are all handled.
• Split by delimiter — the delimiter is literal text, so a full stop means a full stop and a pipe means a pipe - no pattern syntax to escape. Multi-character delimiters work, and \t, \n and \\ are understood so a tab is typeable in a one-line box. If the delimiter does not occur in the text the tool says so instead of handing back the input as though it had done something.

The line-by-line option

Any mode except Line order can be applied within each line instead of across the whole document. Reversing all characters line by line turns every line backwards while keeping the lines in order, which is what most people actually want from a multi-line paste and what a whole-document reversal does not give them. It is switched off in Line order mode because reversing the lines within a single line does nothing, and a control that does nothing is worse than one that is not there.

Mirror does two things, and doing only the first is the bug in most "flip text" generators: it reverses the text, then replaces each character with the glyph that looks like it has been rotated 180 degrees. Rotate without reversing and the result still reads backwards when you turn the screen over.

• Paired characters flip to their partners: an opening bracket becomes a closing one, because the reversal has already swapped which end they sit at.
• The table is its own inverse, so mirroring twice returns exactly what you started with.
• It covers the Latin alphabet, digits and common punctuation. If most of your text has no rotated form - Chinese, Japanese, Korean, Arabic - the tool says so rather than handing back text that has only been reversed.
• Two characters deliberately differ from the usual flip tables, which reach for a Hebrew and an Arabic glyph. Both are strong right-to-left characters, and a single one of them reorders the whole visual line it lands in.

A reverser is a small tool with a surprising number of real jobs, and the mode you want differs for each.

• Reading a log newest-first — Line order, on a .log or .txt dropped straight onto the page. No command line, and it all runs on your device, not our servers.
• Testing text handling — reverse by code units, then by characters, and compare. If your own code produces the first, this is the tool that shows you what your users will see.
• Puzzles, riddles and games — All characters, with the palindrome badge under the result telling you whether the text reads the same both ways.
• Creative and social posts — Mirror, for text that reads correctly upside down, and Letters in each word for a readable-but-wrong effect.
• Reordering data — Split by delimiter for CSV values, path segments or any list joined by a character you can type.
• Language learning and teaching — Letters in each word with punctuation preserved, to demonstrate spelling without destroying the sentence around it.

The free plan handles one document of up to 25 MB, with every mode and every option included - the plan limits scale, never capability. Pro raises the size to whatever your device can handle and lets you queue up to 20 documents and take the reversed set away as a single archive.

• Below 200,000 characters the result updates as you type. Above it, a button applies the change, so a large document never makes typing stutter.
• Reversing runs off the main thread, so the page stays responsive and long runs can be cancelled.
• Only the first 50,000 characters of a result are drawn on screen; Copy and Download always give you the whole thing.
• Text files are decoded from UTF-8, UTF-16 and Latin-1, with or without a byte-order mark. A file that is not text at all is refused with a reason rather than turned into replacement characters.
• Nothing is uploaded, and the tool keeps working with the network disconnected once the page has loaded.

25 MB reverses in about 0.4 seconds on a modern laptop, on your device, with the page still responsive throughout.

For longer workflows, pair this with the text splitter to break a large log into pieces first, the text sorter to order lines before flipping them, and the word counter to check length afterwards. Text splitter · Text sorter · Word counter

• Case inverter

  Swap upper and lower case across a whole document.

• Case converter

  Title case, camel case, snake case and six more.

• ROT13 and Caesar cipher

  Shift letters by any amount, encode and decode.

• Word counter

  Words, characters, sentences and reading time.

• Text sorter

  Sort lines A-Z, by length, or numerically.

• Text splitter

  Split a long document by delimiter, chunk or line count.

• Text deduplicator

  Remove duplicate lines, words or paragraphs.

• Text trimmer

  Strip extra spaces and empty lines.