Repeated spaces
Runs of ordinary U+0020 spaces become one. This does not collapse tabs unless tab conversion is enabled.
Make text spacing consistent
Whitespace Normalizer
Normalize spaces, tabs, no-break spaces, line endings and blank lines with rules you can see. Your original text stays untouched beside a separate result.
- Rule-based cleanup
- Visible whitespace preview
- Zero-width controls preserved
Paste text with inconsistent spacing
Choose a preset, then adjust any individual rule.
Up to 10,000 Unicode code points
Normalization happens in your browser. Text is not uploaded, stored or read from your clipboard automatically. How tool text is handled
NORMALIZED COPY
See every category of change.
Whitespace made visible
· space ⍽ NBSP → tab ↵ line break
SEE IT IN PRACTICE
Follow the spaces through each rule.
Rules run in this order: line endings → NBSP conversion → tab expansion → repeated-space collapse → line-edge trimming → blank-line handling. So a tab set to two spaces can still end as one space when collapse is on. To keep two, turn collapse off; to keep leading indentation, also turn line-edge trimming off.
The first example matches the worked example below. The second uses Line endings only on Item[TAB]Qty and Pen[TAB]2; its output keeps both tabs and the line break.
Choose settings for the text you actually have
| Input and goal | Starting settings | Check before copying |
|---|---|---|
| Prose with accidental NBSP and extra spaces | Clean copied text | NBSP conversion permits wrapping; keep it off where words must stay together. |
| Tab-separated spreadsheet rows | Line endings only; keep tab conversion, collapse and edge trimming off | Tabs may separate columns. Replacing them changes the data structure. |
| Indented code | Line endings only; change individual rules only if required | The tool does not parse code or know which indentation is meaningful. |
| PDF text with a line break after every line | Inspect first | This tool retains non-empty line boundaries. It does not join wrapped lines, repair hyphenated words or reconstruct PDF columns. |
WHAT EACH RULE DOES
Clean spacing without flattening every kind of invisible text.
A text spacing normalizer should transform only the classes you select. It should not silently strip all Unicode formatting.
Tabs
Each U+0009 tab becomes the exact number of regular spaces you choose. Visual alignment can still vary in proportional fonts.
No-Break Space
U+00A0 can be converted to U+0020 when copied web or document text breaks search, matching or wrapping.
Blank lines and edges
Trim line-edge spaces and choose whether to keep all blank lines, keep one between blocks, or remove them.
BOUNDARY WITH HIDDEN CHARACTERS
Zero-width and language controls are preserved.
This tool does not remove U+200B, U+200C, U+200D, bidirectional controls, Hangul Filler or Braille Pattern Blank. That separation prevents an ordinary spacing cleanup from changing scripts or emoji unexpectedly.
FOUR STEPS
Normalize spacing without guessing.
Every rule stays visible, the original text is never replaced, and the change summary names the rule that did the work.
- Paste a copy of the text that already has the wrong spacing.
Paste from the document, spreadsheet export or message where the problem appears. Nothing is uploaded, and the source file is not touched.
- Start from the preset that matches the destination.
“Clean copied text” handles regular spaces, NBSP, tabs, line endings and blank lines. “Keep paragraph spacing” keeps every blank line. “Line endings only” converts CRLF and CR to LF and changes nothing else.
- Turn single rules on or off instead of repairing afterwards.
If a rule would damage the text — tab conversion inside code, collapsed runs inside aligned columns — leave it off. The rule list is deliberately short.
- Compare the visible whitespace, then copy the result.
The result panel shows one count per change category plus a before/after whitespace map. Only the copy button writes a result you can paste elsewhere.
WORKED EXAMPLE
What one run actually reports.
Text copied from a web page often arrives with an NBSP between words, tabs from a table and trailing spaces at the end of lines. Using the notation from the whitespace map:
Pasted text
Report⍽⍽2026→Total·· ··Quarter·1·· ↵ ↵ ··Quarter·2
· regular space ⍽ NBSP → tab ↵ line break
After “Clean copied text”
Report·2026·Total Quarter·1 ↵ Quarter·2
Two NBSP become spaces, the tab becomes two spaces and then collapses to one, other repeated spaces collapse, line edges are trimmed and the double blank line becomes one.
The summary is per rule, not a single “cleaned” badge.
You get separate numbers for line endings, NBSP, tabs, collapsed spaces, trimmed edge spaces and removed blank lines. If a number is larger than expected, switch the preset instead of editing the result by hand.
WHEN THIS IS THE RIGHT TOOL
Spacing, not hidden characters, is the problem.
Text pasted from a PDF, email or web page
Copied paragraphs mix regular spaces, NBSP and tabs. Converting NBSP to U+0020 and collapsing runs makes the paragraph behave like the rest of your document.
Values that fail an exact match
A trailing space or an NBSP inside a copied part number makes two identical-looking values compare as different in a spreadsheet, a filter or an import.
Lines with accidental indentation
Line-edge trimming removes the stray leading space that turns a list entry, a heading or a command into an invalid value.
Drafts that need a predictable paragraph rhythm
“Keep one” blank line between blocks gives consistent paragraph breaks before text is pasted into a form, a CMS field or an email template.
LIMITS
What this normalizer will not fix.
A text that looks empty but is not
If a field still behaves as though it contains something after normalization, the cause is probably a zero-width or blank-looking character. Check the code points before cleaning further.
Spacing you actually want
Alignment built from repeated spaces is fragile in proportional fonts, but the tool will not decide for you. Leave “collapse repeated spaces” off when the repeats are intentional.
Meaning carried by tabs
In source code, build files and tab-separated exports a tab is data. Converting tabs to spaces can change behaviour, which is why tab conversion is an explicit rule instead of automatic.
Unicode form, case and punctuation
Different Unicode representations of the same accented letter, curly versus straight quotes and letter case are outside this tool. It only applies the whitespace rules you select.
QUESTIONS
Whitespace normalizer FAQ
Is this the same as “remove all spaces”?
No. Removing every space joins words together. Normalization keeps the single spaces that separate words and changes only selected classes: runs of spaces, tabs, NBSP, line-edge spaces, line endings and blank lines.
Does it delete zero-width characters such as U+200B?
No. Zero-width, joining and bidirectional controls are preserved here, because removing them can change line breaking, script shaping or emoji. Use the invisible character remover when deletion is the actual task.
Why did my tab become two spaces instead of one?
Tab width is a choice, not a fact about the character. A tab is a single code point with no fixed visual width, so the “Clean copied text” preset expands it to two spaces and the preset label states that. Change the tab width if your destination expects four or eight.
Can normalization change the meaning of my text?
Yes, when whitespace is data. Converted tabs or removed blank lines inside code, indentation-sensitive formats or aligned columns change the text. Review the before/after panel and leave the relevant rule off when the layout carries meaning.
Why does the result still contain invisible characters?
Because this tool only touches spacing characters. If U+200B, U+200C, U+200D, U+2060, U+FEFF, U+2800 or U+3164 are present, they survive normalization on purpose. The checker shows whether they are there.
How much text can I paste, and does it leave my browser?
Up to 10,000 Unicode code points; longer input stops with an explanation instead of a partial result. All processing happens in this page, and the text is not uploaded, stored or read from your clipboard. See how tool text is handled.