01
Start with NFC and compare the four checks
NFC is selected by default, but all four forms are checked together. An accent may look unchanged while its code point sequence changes. The summary calls out forms that already match the input.
Unicode and Internationalization
Unicode normalizersee what really changes
Compare all four forms against the original, then inspect the selected result before using it.
Unicode normalizer
Original text
Selected normalized result
Preview shows at most 4,000 characters; copy and download include the complete result.
Text is processed in this browser and is not uploaded.
Up to 10,000 UTF-16 units. Oversized input is rejected without truncation.
Normalization uses this browser's Unicode implementation; no fixed Unicode version is claimed.
Enter text or choose an example.
When two names look identical but a search or equality check fails, the stored Unicode sequence may differ. This Unicode normalizer makes a precomposed accent and its decomposed equivalent visible instead of guessing from the rendered glyph.
The Unicode normalizer also exposes compatibility mappings that can alter presentation or meaning. Start with NFC for ordinary text, examine the changed code points and bytes, and apply a result to the input only after you review it.
Why use this Unicode normalizer?
See the sequence behind a familiar looking string.
Compare four forms from the same source
A Unicode normalization checker should show which forms already match the original. Here NFC, NFD, NFKC and NFKD run independently on the same input; selecting one reveals its exact output without chaining transformations.
Catch compatibility changes before replacement
A ligature or fullwidth letter can become ordinary letters under NFKC or NFKD. The Unicode normalizer marks changed ranges and shows U+ values and UTF-8 bytes, while the input stays intact until you confirm a preview.
Choose a form without losing the source
02
Use line mode only for independent records
Whole text retains its line endings. Line mode treats each logical line as a separate record, preserves empty rows and joins the results with LF. It accepts 500 records; use whole text when exact original separators matter.
03
Review the difference before exporting
The highlighted range is a sequence change, not a one-character mapping. Read U+ values and UTF-8 bytes, then copy or download the full result. Turn off automatic checking to run manually with Ctrl/Cmd+Enter.
What the comparison reveals
Four independent normalization checks
An NFC converter and NFD converter are useful for different storage expectations. The summary shows whether each form would change the original and reports its code point and byte totals.
Changed ranges with two representations
Inspect changed code point sequences as U+ values and their UTF-8 bytes. Composition and decomposition can involve several input and output code points, so the tool groups edits into ranges.
Controlled replacement and full export
Copy and download use the complete selected result even when the visual preview is shortened. Applying to input requires a before-and-after preview, and one undo restores both input and prior results.
Boundaries of the conversion
- Compatibility forms may lose distinctions
- NFKC normalization and NFKD can fold width, ligatures and superscripts. This is not a reversible formatting repair. Compare before applying when those distinctions are meaningful.
- No accent stripping or case folding
- NFD separates a base letter from its combining mark but does not remove the mark. The Unicode normalizer never silently turns text into ASCII, changes letter case or performs NFKC_Casefold.
- No confusable or invisible-character repair
- Latin and Cyrillic lookalikes, zero-width characters and mixed scripts are not fixed by these four forms. Use a character inspector for those questions instead of treating normalization as a security check.
- Bounded pasted text, no file parsing
- The tool accepts at most 10,000 UTF-16 units and 500 records in line mode. It rejects isolated surrogates and oversize results. It does not upload files, parse CSV or JSON, or preserve CRLF separators in line mode.
Unicode normalizer questions
How do I normalize text to NFC or NFD?
Paste text and select NFC or NFD. The Unicode normalizer checks all forms against the original and displays the selected result. NFC composes where possible; NFD decomposes canonically. Copy the result or preview a replacement before applying it.
What is the difference between NFC, NFD, NFKC, and NFKD?
NFC and NFD use canonical equivalence, with composition or decomposition. NFKC and NFKD also apply compatibility mappings, which may remove width, ligature or superscript distinctions. The four outputs are calculated independently.
Why can identical-looking text compare differently?
A glyph such as é may be stored as one precomposed code point or as e plus a combining accent. Display can look the same while code points and UTF-8 bytes differ. The highlighted comparison makes that difference visible.
Is my text uploaded?
No. Checking, diffing, copying and download creation run in your browser. The page does not send the input to a conversion server.
What does this Unicode normalizer not support?
It does not strip accents, fold case, detect spoofing, upload files or use a fixed Unicode normalization data version. Line mode accepts pasted records rather than a CSV or JSON file.
Still seeing an unexpected sequence?
Try the Unicode normalizer on your text
Start with a small fragment that fails a comparison, inspect all four forms, and keep the original until you choose a result.