Same letters, different middle
Each line contains A, one character and B.
Choose by behavior
Characters that look empty can behave very differently. Compare visible width, line breaking and Unicode classification before you copy one.
Copy Hangul Filler and see its full character reference.
U+2800Try a wider blank cellBraille Pattern Blank has font-dependent width and may render differently.
U+200BAdd no visible widthCopy Zero Width Space and see its line-wrap demonstration.
U+00A0Keep words togetherNo-Break Space usually looks ordinary but prevents a line break.
“Invisible,” “blank” and “whitespace” are useful everyday descriptions, but they are not interchangeable Unicode properties.
| Character | Unicode property | Width & breaking | Use when | Action |
|---|---|---|---|---|
| Regular SpaceU+0020 · Space Separator | White_Space: Yes | Visible gap; normal line-break opportunity. | You need ordinary word spacing. | Baseline |
| No-Break SpaceU+00A0 · Space Separator | White_Space: Yes | Usually a visible gap; prevents a break between adjacent text. | A short phrase or value must stay together. | Use U+00A0 |
| Zero Width SpaceU+200B · Format | White_Space: No | No visible width; creates a break opportunity. | You need an invisible word-break hint. | Use U+200B |
| Braille Pattern BlankU+2800 · Other Symbol | White_Space: No | Blank-looking Braille cell; width depends on the font. | You want an alternate blank-looking character. | Use U+2800 |
| Hangul FillerU+3164 · Other Letter | White_Space: No | Often blank-looking and width-bearing; display depends on the font. | You want text that appears blank in a compatible field. | Use U+3164 |
A character can look blank without being Unicode whitespace. Fonts, input rules, trimming and validation can still change or reject it after pasting.
Boundary marks and counts make otherwise invisible differences easier to inspect.
Each line contains A, one character and B.
A regular space can break; U+00A0 keeps the adjacent words together when line breaking is supported.
empty0U+00201U+200B1Names, general categories and White_Space values come from the Unicode Character Database dated August 7, 2026. Display and acceptance still depend on the target font and app.
Answer them in order. The first answer that matches your goal already names the character to try.
If the answer is yes, the character must have width. Choose U+0020 for ordinary spacing, U+00A0 when the gap must resist a line break, or U+2003 Em Space when the gap should be wider than an ordinary space.
Then the character must have no width. U+200B Zero Width Space gives a possible break without a visible mark, and U+2060 Word Joiner does the opposite: it blocks a break without adding width.
Then choose a blank-looking letter or symbol rather than a space: U+3164 Hangul Filler is a letter, and U+2800 Braille Pattern Blank is a symbol. Neither is Unicode whitespace, so trimming rules may treat them differently.
Copy, paste, save, reopen, and check the stored value with the checker. Every platform decides for itself whether to keep, normalize or reject a character.
Start with a blank-looking, width-bearing character such as U+3164 or U+2800 rather than a space, because spaces are commonly trimmed. Then verify what the platform stored.
Use U+200B when the separator must not add width, or a width-bearing blank when the layout needs an actual gap between blocks of text.
Use U+00A0 between the words or around a number, for example between a quantity and its unit.
Use U+200B where the browser may break the string. Nothing visible changes, but the line can now wrap at that point.
Use no extra character at all. Any blank-looking or zero-width code point makes two otherwise identical values different, which is exactly what breaks an exact match.
Do not guess from appearance. Check the code points first, then decide whether to keep, replace or remove them.
Whitespace is an official Unicode property. Blank and invisible are everyday descriptions of appearance. U+3164 and U+2800 look blank but are not whitespace, and U+200B has no visible width while also not being whitespace, so the words cannot be used interchangeably when you predict app behaviour.
There is no single answer, because reliability means “what the destination keeps”. U+00A0 and U+0020 are supported almost everywhere as spacing; U+200B, U+2800 and U+3164 depend on font, input validation and normalization rules.
Fonts. A character with no glyph in the active font is rendered through font fallback, which can produce a box, a dot, different spacing or nothing at all. The code point is unchanged; only the rendering differs.
Yes for width-bearing characters such as U+00A0, U+2800 and U+3164. Repeating U+200B does not create width, because it has none; several in a row only add more break opportunities.
Yes. Each one is a code point, so a counter includes it even when the screen shows nothing. That is why a value can look empty and still be rejected for being too long.
Open the character directory for the full list with per-character copy actions, or use the main generator to produce an exact amount of one character.