Text Scrambler
Shuffle characters or words with a reproducible seed and adjustable intensity.
Settings
The same text, seed, and settings always produce the same result.
How Text Scrambler works
Scrambling permutes the characters or words of your text without adding or removing any. Because it is a permutation, the output always has exactly the same length and the same multiset of symbols as the input — only the order changes.
The shuffle is driven by a seeded pseudo-random number generator. Given the same seed, the same text, and the same settings, the generator emits the same sequence of numbers and therefore produces byte-for-byte the same scrambled output. That reproducibility is the point: you can share a seed instead of the result and let someone else regenerate it.
Intensity controls how many positions actually move. At low intensity only a fraction of the characters are swapped, leaving the text recognisable but visibly disordered; at full intensity the permutation touches everything, using a Fisher–Yates shuffle that gives every arrangement equal probability.
Word mode keeps each word intact and shuffles the words themselves, which preserves spelling and readability of individual tokens while destroying sentence structure. Character mode operates within the text as a whole.
How to use this scrambler
Paste your text
Drop in the text you want to scramble. Length is not limited by the tool — everything runs locally.
Pick a mode
Character mode shuffles individual letters; word mode keeps words whole and reorders them.
Set the seed and intensity
Leave the seed blank for a fresh random result, or enter one to make the output reproducible. Raise intensity for a more thorough shuffle.
Copy the result
Copy the scrambled text, and note the seed if you will need to reproduce exactly this output later.
Worked examples
Reproducible test fixture
- Given
- "the quick brown fox" in word mode with seed 42
- Result
- "brown the fox quick"
Re-running with seed 42 gives the same arrangement every time, so a test asserting on that string stays stable across runs.
Low-intensity character shuffle
- Given
- "Configuration" in character mode at 25% intensity
- Result
- "Cnofigaruiton"
Only a quarter of the positions moved, so the word remains recognisable — useful for typo-tolerance and fuzzy-search demos.
Anonymising sample copy
- Given
- A paragraph of real customer feedback, word mode, full intensity
- Result
- The same words with no recoverable sentences
Good enough for a layout mockup. Not anonymisation — rare words like a company name still identify the source.
When to use it
- Producing deterministic placeholder text for UI mockups without shipping real copy.
- Generating stable fixtures for tests of search ranking, fuzzy matching, or spell correction.
- Building word-jumble and anagram puzzles with a shareable seed.
- Stress-testing text rendering, wrapping, and truncation with realistic character distributions.
- Demonstrating how much word order matters to a tokeniser or language model.
Things to watch out for
- Scrambling is not encryption. It is a permutation with no key and no security properties — frequency analysis recovers a great deal, and short texts are trivially unscrambled by hand.
- The multiset of characters is preserved exactly, so anything identifiable that survives as a single token — an email address in word mode, a rare proper noun — still leaks.
- The same seed only reproduces output for identical input and identical settings. Change one character of the source and the result diverges completely.
- Character mode moves whitespace and punctuation like any other character, which is why output can gain odd spacing.
Frequently asked questions
Can scrambled text be unscrambled?
Yes, given the seed and settings the permutation is fully reversible, and even without them short texts fall to simple analysis. Treat this as a formatting tool, never as a way to protect information.
What does the seed actually do?
It initialises the pseudo-random generator that chooses the swaps. Identical seed plus identical input plus identical settings always produce identical output, which is what makes results shareable and testable.
Is this safe for anonymising real data?
No. Every character survives the shuffle, so unique tokens — names, domains, order numbers — remain present and often intact. Use proper redaction or synthetic data generation for anything sensitive.
What is the difference between character and word mode?
Character mode shuffles letters across the whole text, destroying spelling. Word mode keeps each word intact and reorders the words, which preserves vocabulary and readability while destroying grammar.
Does intensity below 100% weaken reproducibility?
No. Intensity changes how many positions the algorithm swaps, but the choice of which ones still comes from the seeded generator, so any intensity setting reproduces exactly with the same seed.
Related Developers tools
All developer tools- JWT Generator Build and sign HS256 JSON Web Tokens from a header, payload, and secret.
- JWT Decoder Inspect a token’s header and payload, with optional signature verification.
- Base64 Encode / Decode Encode and decode Base64 for text and files, with URL-safe support.
- Encrypt / Decrypt Text Encrypt and decrypt text with a passphrase (AES-GCM/CBC) or an RSA public key — output as Base64.
- Hash Generator Compute MD5, SHA-1, SHA-256, and SHA-512 digests, or Argon2 password hashes with a pepper.
- JSON Formatter Pretty-print, minify, and validate JSON with precise error positions and a tree view.