ROT13 & Caesar Cipher Solver, Encoder and Decoder

Decode, encode and brute-force Caesar ciphers, ROT13, ROT5, ROT18, ROT47 and progressive shifts in one place. Solve tries every shift and ranks the likely plaintext. Decode applies a shift or variant you already know. Encode rotates plain text forward, so you can build a test cipher or check a round trip.

It also reads rotated CTF flag prefixes: if the text contains something like synt{, the shift that turns it into flag{ is shown first. You can rotate the whole text, only the part inside the outermost braces, a selection, or each line on its own (up to 500 lines), which suits log files and puzzles where only part of the text is enciphered. Input is limited to 64 KB.

Candidates are scored with English letter statistics by default, and German, French, Spanish, Italian, Portuguese or Russian when you choose one. The tool reports a confidence level, not a percentage, and with fewer than 8 letters it never names a winner. Results can be copied, exported as JSON or CSV, or sent to the next tool. Everything runs in your browser.

A Caesar cipher replaces every letter with the letter a fixed number of places further along the alphabet, so there are only 25 useful keys. Brute force simply tries them all: this tool decrypts the text with every shift from 0 to 25 and scores each result by how much it looks like the language you picked — how close its letter frequencies are to normal English, how many of its letter pairs are common ones such as th and er, and how much of it is made of common words. The highest score is the likely plaintext.

That works well for sentences and fails for tiny inputs, which is why the tool reports a Solver confidence instead of a percentage. With 20 or more letters and a clear gap over the runner-up it says High. With 8 to 19 letters the best it will say is Low, because a handful of letters can look like English by chance. Under 8 letters it says Unreliable and does not highlight a winner at all.

ROT13 is a Caesar shift of 13. Because the alphabet has 26 letters, applying it twice returns the original, so decoding and encoding are the same operation. ROT5 does the same for the ten digits, ROT18 combines ROT13 for letters with ROT5 for digits, and ROT47 rotates the 94 printable ASCII characters from ! to ~ by 47 places, which is why it also scrambles punctuation and digits.

Pick the variant from what the challenge text contains. A flag with digits that came out as ROT13 will still show its digits unchanged, and the tool says so in the result notes so you can switch to ROT18 or ROT47 instead of wondering why part of the text looks wrong.

Frequency analysis needs text, but a flag is often shorter than twenty letters. What a flag does have is a known beginning: flag{, ctf{, htb{, picoctf{. If the ciphertext starts synt{, then s→f, y→l, n→a and t→g are all the same distance apart — 13 places — and four letters agreeing on one shift is strong evidence even though four letters could never be ranked statistically.

The prefix finder looks for any run of letters directly before a brace, compares it with every known prefix of the same length, and proposes a shift only when every letter agrees. It also finds the token inside longer text, and you can add your own prefixes (for example thm) so it works for any event's flag format. The result appears above the ranked candidates, with the decoded flag so you can confirm it before using it.

If none of the 26 results reads as language, the text is probably not a single-shift cipher. Short text, text with only the content of the braces rotated, and text in another language all look the same to a frequency table. When there is enough text the tool says so plainly and shows what to try next: an Index of Coincidence well below the language's normal value is a clue for a polyalphabetic cipher such as Vigenère, an alphabet-reversing Atbash output that reads as text points to Atbash, and mostly non-letter characters suggest ROT47 or an XOR-style cipher. Each suggestion shows the numbers behind it and is only a hint, never a verdict.

Challenge authors often rotate only what is inside the braces and leave the prefix alone, so rotating the whole text turns flag{ into garbage. Choose Inside { } to rotate only the contents of the outermost balanced pair of braces; nested or extra groups are reported, not silently changed. Selection rotates only the range you highlighted in the box, and Per line treats every line as its own puzzle with its own best shift and its own confidence, which is the right choice when each line was encrypted separately. Per-line mode handles the first 500 lines.

A progressive Caesar cipher changes the shift with every letter: the key for the nth letter (counting from 0, letters only) is start + n × step, so attack with start 1 and step +1 becomes bvwehq. With a known start and step, Decode and Encode apply it directly; in Solve mode the tool tries all 676 start/step combinations over the first 4,096 characters and ranks them.

A custom alphabet lets you rotate any ordered list of symbols — the Russian or Greek alphabet, letters plus digits, a CTF-specific symbol set. Symbols must be unique, characters outside the alphabet pass through unchanged, and case is matched loosely when every symbol has a distinct lower-case form. German, French, Spanish, Italian and Portuguese data is loaded only when you choose that language, so the English page stays light.