Normalize Text
Clean and normalize text in one pass — apply a Unicode form, fold case, strip accents, collapse whitespace, and straighten fancy punctuation. Runs in your browser; nothing is uploaded.
About this tool
Text Normalizer cleans and canonicalizes text in one configurable pass — ideal for building search keys, deduplicating records, preparing NLP input, or just tidying text pasted from PDFs, chat apps, and word processors. Everything runs locally in your browser through WebAssembly; your text is never uploaded.
The pipeline runs in a fixed, predictable order:
- Straighten punctuation — fold curly quotes, guillemets, en/em dashes, the
ellipsis glyph, and prime marks to plain ASCII (
",',-,...). - Unicode form — apply NFC, NFD, NFKC, or NFKD. NFC is the web-recommended canonical form; NFKC additionally flattens fancy fonts — ligatures (fi → fi), full-width and circled/styled letters (𝐇𝐞𝐥𝐥𝐨 → Hello), and fractions.
- Strip accents — decompose letters and drop combining marks (café → cafe, naïve → naive).
- Case — lowercase, UPPERCASE, or case-fold. Case fold is for caseless matching: like lowercase, but it also maps ß → ss so STRASSE and straße compare equal.
- Whitespace — collapse every run of whitespace to a single space (the default), trim the ends only, or collapse each line while keeping line breaks.
Whitespace runs last so it tidies anything the earlier steps introduce. Every option is independent — turn on only what you need, and the defaults (NFC + collapse whitespace) already give clean, web-safe text.
Worked example
Input (NLP-prep preset — NFKC, case fold, strip accents, collapse, straighten punctuation):
Café “RÉSUMÉ”…
Output:
cafe "resume"...
FAQ
What is Unicode normalization, and which form should I pick?
Unicode often has several ways to encode the same visible text — for example an
accented é can be one code point or an e plus a combining accent. Normalization
rewrites text to a single canonical representation so comparisons and searches
behave. Use NFC (the default) for general web text, NFKC when you also want
to flatten ligatures and fancy-font letters to their plain equivalents, and the
NFD/NFKD decomposed forms when a downstream tool expects separated base letters
and marks.
What is the difference between lowercase and case fold?
Lowercase applies the standard Unicode lowercase mapping. Case fold is designed
for caseless matching rather than display: it lowercases and additionally expands
a few characters that have no single lowercase form — most notably the German
ß, which folds to ss. Use case fold when you want STRASSE, Straße, and
strasse to all compare equal; use plain lowercase when you just want readable
lower-case text.
Does this delete punctuation or numbers like an NLP cleaner?
No. Text Normalizer normalizes — it does not destroy content. Turning on Straighten fancy punctuation converts typographic characters (curly quotes, em dashes, the ellipsis glyph) to their ASCII equivalents, but it never removes punctuation or digits. If you need to strip punctuation or numbers entirely, that is a separate destructive cleaning step outside this tool's scope.
Is my text uploaded anywhere?
No. The tool is a WebAssembly module that runs entirely in your browser. Your text never leaves your device — there is no server round-trip, so it is safe for private or sensitive content.
Developer & Automation Access
Run it from the terminal
Same engine as this page, headless — via the gizza CLI:
gizza tool text-normalizer "Paste text to normalize…"New to the CLI? Get gizza →
Open it by URL
Pre-fill and auto-run this tool with query parameters — the names match the API/CLI:
https://gizza.ai/tools/text-normalizer/?text=Paste%20text%20to%20normalize%E2%80%A6&form=nfc&case=none&strip_accents=true&whitespace=collapse&punctuation=trueMachine-readable descriptor: tool.json — title + parameters JSON Schema for agents.