90 of the smallest visible characters in five groups, each ordered smallest first as our screen drew them. The smallest symbol of all was ་, the Tibetan tsheg, a delimiter. Some apps strip or block these characters, and screen readers may read them badly.
Read the full guide
How the order was made
Unicode gives a character no size: the font decides. So we drew all 90 at one size in the system fonts of one computer, counted the dark pixels and checked each group by eye. Dots opens with ་, Short dashes with ˗ and The smallest letters with ʳ. Another font will shuffle neighbours. Unicode does name a few shapes tiny, but they drew as empty boxes here and were left out.
The same mark at four heights
A small mark also has a place. Search raised for the 35 that sit near the top of a letter, such as ˙, and below for the 14 that hang under the line, such as ˎ. The eight Braille cells in Dots put one dot in eight places.
Tips
Search keyboard for the 5 you can already type: they are here as yardsticks for the rest.
Text normalised to NFKC loses 30 of the 90: ﹒ becomes an ordinary full stop and ˣ an x.
Search wide for 5 marks that arrive in a full square cell, such as ﹒: tiny ink, a big gap.
Questions and answers
What is the smallest symbol you can copy and paste?
On our screen it was ་, the Tibetan tsheg at U+0F0B, with ﹒, ॱ and ˙ close behind. No character has a fixed size, so another font may crown a different one. The plain full stop came sixth among the dots.
What is the smallest letter?
Of the 16 we drew, ʳ left the least ink, then ˡ. Three look like apostrophes and are letters: the saltillo ꞌ, the modifier letter apostrophe ʼ and ʻ, the Hawaiian ʻokina. The other 13 turn into ordinary letters under NFKC.
Is there a symbol smaller than a dot?
Only one with no ink at all. Unicode has blank characters, some of them with no width, and copying those is a different job: the invisible text generator draws each one between two lines so that its width shows.
⠁Braille dot 1U+2801 Braille pattern dots-1. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 1 raised: top left. One dot, eight possible places, all in a cell of the same width.
⠂Braille dot 2U+2802 Braille pattern dots-2. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 2 raised: middle left. One dot, eight possible places, all in a cell of the same width.
⠄Braille dot 3U+2804 Braille pattern dots-3. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 3 raised: lower left. One dot, eight possible places, all in a cell of the same width.
⠈Braille dot 4U+2808 Braille pattern dots-4. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 4 raised: top right. One dot, eight possible places, all in a cell of the same width.
⠐Braille dot 5U+2810 Braille pattern dots-5. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 5 raised: middle right. One dot, eight possible places, all in a cell of the same width.
Show all 90 characters (85 more)
Dots, continued
⠠Braille dot 6U+2820 Braille pattern dots-6. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 6 raised: lower right. One dot, eight possible places, all in a cell of the same width.
⡀Braille dot 7U+2840 Braille pattern dots-7. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 7 raised: bottom left, the extra row of eight-dot Braille. One dot, eight possible places, all in a cell of the same width.
⢀Braille dot 8U+2880 Braille pattern dots-8. Braille Patterns block, Unicode 3.0. A Braille cell with only dot 8 raised: bottom right, the extra row of eight-dot Braille. One dot, eight possible places, all in a cell of the same width.
•BulletU+2022 Bullet. General Punctuation block, Unicode 1.1. The Unicode names list also calls it black small circle. It is the largest dot here, kept for scale.
ᐧCanadian syllabics final middle dotU+1427 Canadian syllabics final middle dot. Unified Canadian Aboriginal Syllabics block, Unicode 3.0. A letter, not punctuation. The Unicode names list names Moose Cree, Athapascan, Sayisi and Blackfoot as its users.
ॱDevanagari high spacing dotU+0971 Devanagari sign high spacing dot. Devanagari block, Unicode 5.1. A raised dot from the Devanagari block. Unicode files it as a modifier letter, not as punctuation.
˙Dot aboveU+02D9 Dot above. Spacing Modifier Letters block, Unicode 1.1. The dot of an i with no i under it. The Unicode names list gives one use: the fifth, neutral tone of Mandarin.
⋅Dot operatorU+22C5 Dot operator. Mathematical Operators block, Unicode 1.1. The multiplication dot. The Unicode names list prefers it to the middle dot for that job.
.Full stopU+002E Full stop. Basic Latin block, Unicode 1.1. The one on every keyboard, kept here as the yardstick for the rest.
．Fullwidth full stopU+FF0E Fullwidth full stop. Halfwidth and Fullwidth Forms block, Unicode 1.1. A full stop in a full square cell: Unicode maps it to the ordinary full stop, marked wide.
･Halfwidth katakana middle dotU+FF65 Halfwidth Katakana middle dot. Halfwidth and Fullwidth Forms block, Unicode 1.1. The katakana middle dot squeezed into half a cell: Unicode maps it to that character, marked narrow.
‧Hyphenation pointU+2027 Hyphenation point. General Punctuation block, Unicode 1.1. Dictionaries use it to show where a word may break: the Unicode names list prints dic·tion·ar·ies.
・Katakana middle dotU+30FB Katakana middle dot. Katakana block, Unicode 1.1. The middle dot of the Katakana block. Unicode classes it as wide: a full square cell.
·Middle dotU+00B7 Middle dot. Latin-1 Supplement block, Unicode 1.1. The Unicode names list gives it three more names: midpoint, Georgian comma and Greek ano teleia.
․One dot leaderU+2024 One dot leader. General Punctuation block, Unicode 1.1. One dot of the dotted row that leads across a contents page. The Unicode names list adds that Armenian uses it as a semicolon, the mijaket.
᛫Runic single punctuationU+16EB Runic single punctuation. Runic block, Unicode 3.0. Punctuation from the Runic block: one dot at mid height, a little heavier than the middle dot on our screen.
﹒Small full stopU+FE52 Small full stop. Small Form Variants block, Unicode 1.1. A compatibility variant for the standard CNS 11643, says the Unicode names list. The dot is small but the cell is not: Unicode classes the character as wide, so it arrives in a full square.
༌Tibetan non-breaking tshegU+0F0C Tibetan mark delimiter tsheg bstar. Tibetan block, Unicode 2.0. The tsheg again with one difference: the Unicode names list says it inhibits line breaking, so the line never ends after it.
་Tibetan tshegU+0F0B Tibetan mark intersyllabic tsheg. Tibetan block, Unicode 2.0. The everyday delimiter of Tibetan writing: the Unicode names list calls it the normal tsheg and also spells it tsek. On our screen it left less ink than anything else here. It sits high, level with the top of a letter x.
⸱Word separator middle dotU+2E31 Word separator middle dot. Supplemental Punctuation block, Unicode 5.2. A dot made to separate words; the Unicode names list mentions Avestan and Samaritan. Its cell is narrower than the middle dot's on our screen.
Ticks, commas and accents 19
´Acute accentU+00B4 Acute accent. Latin-1 Supplement block, Unicode 1.1. The mark of é with no letter under it.
'ApostropheU+0027 Apostrophe. Basic Latin block, Unicode 1.1. The straight one on every keyboard, as a yardstick. The Unicode names list prefers U+2019, the curly one, for real apostrophes.
¸CedillaU+00B8 Cedilla. Latin-1 Supplement block, Unicode 1.1. The hook under ç with no letter above it.
˓Centred left half ringU+02D3 Modifier letter centred left half ring. Spacing Modifier Letters block, Unicode 1.1. A phonetic mark: less rounded articulation, says the Unicode names list.
˒Centred right half ringU+02D2 Modifier letter centred right half ring. Spacing Modifier Letters block, Unicode 1.1. More rounded articulation: the partner of the left half ring.
,CommaU+002C Comma. Basic Latin block, Unicode 1.1. On every keyboard: the yardstick for the marks that hang below the line.
`Grave accentU+0060 Grave accent. Basic Latin block, Unicode 1.1. On every keyboard. The Unicode names list also calls it backtick and backquote.
῾Greek dasiaU+1FFE Greek dasia. Greek Extended block, Unicode 1.1. The rough breathing mark: the psili turned the other way.
͵Greek lower numeral signU+0375 Greek lower numeral sign. Greek and Coptic block, Unicode 1.1. Set low before a Greek letter to show that the letter is a number. The Unicode names list calls it aristeri keraia.
᾿Greek psiliU+1FBF Greek psili. Greek Extended block, Unicode 1.1. The smooth breathing mark of ancient Greek, as a character of its own.
⸒HypodiastoleU+2E12 Hypodiastole. Supplemental Punctuation block, Unicode 4.1. A Greek mark that the Unicode names list also spells ypodiastoli. It stands on the line where a comma would hang below it.
ˏLow acute accentU+02CF Modifier letter low acute accent. Spacing Modifier Letters block, Unicode 1.1. Its twin: low-rising tone.
ˎLow grave accentU+02CE Modifier letter low grave accent. Spacing Modifier Letters block, Unicode 1.1. A tone mark set under the line: low-falling tone.
ʹModifier letter primeU+02B9 Modifier letter prime. Spacing Modifier Letters block, Unicode 1.1. A slim raised tick. The Unicode names list gives two uses: stress, and the Cyrillic soft sign in transliteration.
˛OgonekU+02DB Ogonek. Spacing Modifier Letters block, Unicode 1.1. The hook under ą with no letter above it.
′PrimeU+2032 Prime. General Punctuation block, Unicode 1.1. Minutes and feet, says the Unicode names list.
﹐Small commaU+FE50 Small comma. Small Form Variants block, Unicode 1.1. From the same CNS 11643 set as the small full stop. Unicode classes it as wide, so the cell is a full square.
ˈVertical line, highU+02C8 Modifier letter vertical line. Spacing Modifier Letters block, Unicode 1.1. The primary stress mark of phonetic writing. The Unicode names list says it goes before the syllable it marks.
ˌVertical line, lowU+02CC Modifier letter low vertical line. Spacing Modifier Letters block, Unicode 1.1. Secondary stress: the same stroke at the foot of the line.
Short dashes 11
⁃Hyphen bulletU+2043 Hyphen bullet. General Punctuation block, Unicode 1.1. A hyphen heavy enough to open a list item.
-Hyphen-minusU+002D Hyphen-minus. Basic Latin block, Unicode 1.1. The dash on every keyboard, as a yardstick. The Unicode names list says it serves generically as hyphen, minus sign and en dash.
ˍLow macronU+02CD Modifier letter low macron. Spacing Modifier Letters block, Unicode 1.1. A tone mark under the line: low level tone.
¯MacronU+00AF Macron. Latin-1 Supplement block, Unicode 1.1. The Unicode names list also calls it overline and APL overbar, and notes that two of them side by side do not join.
⧿MinyU+29FF Miny. Miscellaneous Mathematical Symbols-B block, Unicode 3.2. Its whole Unicode name is Miny. The names list says that Tiny and Miny yield infinitesimal values in combinatorial game theory.
ˉModifier letter macronU+02C9 Modifier letter macron. Spacing Modifier Letters block, Unicode 1.1. High level tone. The Unicode names list adds that it is the first tone of Mandarin.
˗Modifier letter minus signU+02D7 Modifier letter minus sign. Spacing Modifier Letters block, Unicode 1.1. A phonetic mark for retracted articulation, and the shortest dash here on our screen.
꞊Short equals signU+A78A Modifier letter short equals sign. Latin Extended-D block, Unicode 5.1. Two short dashes. The Unicode names list says it is a tone letter for Budu.
﹣Small hyphen-minusU+FE63 Small hyphen-minus. Small Form Variants block, Unicode 1.1. Another CNS 11643 compatibility variant. Unicode classes it as wide: a short dash in a full square cell.
₋Subscript minusU+208B Subscript minus. Superscripts and Subscripts block, Unicode 1.1. A minus sign at subscript size, resting near the baseline.
⁻Superscript minusU+207B Superscript minus. Superscripts and Subscripts block, Unicode 1.1. The same minus sign lifted to superscript height.
Small shapes 19
⬩Black small diamondU+2B29 Black small diamond. Miscellaneous Symbols and Arrows block, Unicode 5.1. A solid diamond at the small size.
▪︎Black small squareU+25AA Black small square. Geometric Shapes block, Unicode 1.1. The Unicode names list also calls it square bullet. Has an emoji form too, so it is copied with U+FE0E, the selector that keeps it as plain text.
⬝Black very small squareU+2B1D Black very small square. Miscellaneous Symbols and Arrows block, Unicode 5.1. The solid one of the pair that Unicode names very small.
˟Cross accentU+02DF Modifier letter cross accent. Spacing Modifier Letters block, Unicode 3.0. A raised x. The Unicode names list gives its use as the Swedish grave accent.
°Degree signU+00B0 Degree sign. Latin-1 Supplement block, Unicode 1.1. The familiar raised ring, kept for scale: the ring above is about half its size on our screen.
⋄Diamond operatorU+22C4 Diamond operator. Mathematical Operators block, Unicode 1.1. A maths operator. On our screen it is the smallest shape with corners; a maths font may draw it larger.
˅Down arrowheadU+02C5 Modifier letter down arrowhead. Spacing Modifier Letters block, Unicode 1.1. Lowered articulation.
ˑHalf triangular colonU+02D1 Modifier letter half triangular colon. Spacing Modifier Letters block, Unicode 1.1. The half-length mark of phonetic writing. The length mark itself is the triangular colon, U+02D0.
˂Left arrowheadU+02C2 Modifier letter left arrowhead. Spacing Modifier Letters block, Unicode 1.1. Fronted articulation.
˳Low ringU+02F3 Modifier letter low ring. Spacing Modifier Letters block, Unicode 4.0. A ring that hangs under the line.
˖Modifier letter plus signU+02D6 Modifier letter plus sign. Spacing Modifier Letters block, Unicode 1.1. A phonetic mark: advanced or fronted articulation. A plus sign that rests on the baseline.
˃Right arrowheadU+02C3 Modifier letter right arrowhead. Spacing Modifier Letters block, Unicode 1.1. Backed articulation.
˚Ring aboveU+02DA Ring above. Spacing Modifier Letters block, Unicode 1.1. The ring of å with no letter under it. Smaller than the degree sign.
⋆Star operatorU+22C6 Star operator. Mathematical Operators block, Unicode 1.1. A maths operator, and the smallest five-pointed star we drew: the character Unicode names white small star left three times the ink.
⧾TinyU+29FE Tiny. Miscellaneous Mathematical Symbols-B block, Unicode 3.2. Its whole Unicode name is Tiny, and it is not: it draws as a plus sign with a bar under it. The names list says that Tiny and Miny yield infinitesimal values in combinatorial game theory.
˄Up arrowheadU+02C4 Modifier letter up arrowhead. Spacing Modifier Letters block, Unicode 1.1. A phonetic mark: raised articulation.
◦White bulletU+25E6 White bullet. Geometric Shapes block, Unicode 1.1. A bullet in outline: a ring at mid height.
▫︎White small squareU+25AB White small square. Geometric Shapes block, Unicode 1.1. One size up from very small. Has an emoji form too, so it is copied with U+FE0E, the selector that keeps it as plain text.
⬞White very small squareU+2B1E White very small square. Miscellaneous Symbols and Arrows block, Unicode 5.1. Very small is part of its Unicode name. The outline of the two.
The smallest letters 16
ʼModifier letter apostropheU+02BC Modifier letter apostrophe. Spacing Modifier Letters block, Unicode 1.1. Looks like a curly apostrophe and is a letter: the Unicode names list says many languages use it as a letter of their alphabets.
ʻʻOkinaU+02BB Modifier letter turned comma. Spacing Modifier Letters block, Unicode 1.1. A letter of Hawaiian, the ʻokina. The Unicode names list says some Polynesian orthographies use it for the glottal stop.
ꞌSaltilloU+A78C Latin small letter saltillo. Latin Extended-D block, Unicode 5.1. A lowercase letter that looks like a straight apostrophe. The Unicode names list says saltillos stand for the glottal stop and that the lowercase is widely used in many languages in Mexico.
ᵢSubscript iU+1D62 Latin subscript small letter i. Phonetic Extensions block, Unicode 4.0. A lowered i.
ⱼSubscript jU+2C7C Latin subscript small letter j. Latin Extended-C block, Unicode 5.1. A lowered j. It came in Unicode 5.1, later than most of the row.
ₗSubscript lU+2097 Latin subscript small letter l. Superscripts and Subscripts block, Unicode 6.0. The thin stroke again, lowered.
ᵣSubscript rU+1D63 Latin subscript small letter r. Phonetic Extensions block, Unicode 4.0. A lowered r.
ₓSubscript xU+2093 Latin subscript small letter x. Superscripts and Subscripts block, Unicode 4.1. The letter x lowered and shrunk.
ᴵSuperscript capital IU+1D35 Modifier letter capital i. Phonetic Extensions block, Unicode 4.0. A capital I at superscript size.
ⁱSuperscript iU+2071 Superscript Latin small letter i. Superscripts and Subscripts block, Unicode 3.2. A raised i, dot included.
ʲSuperscript jU+02B2 Modifier letter small j. Spacing Modifier Letters block, Unicode 1.1. In phonetic writing it marks palatalization.
ˡSuperscript lU+02E1 Modifier letter small l. Spacing Modifier Letters block, Unicode 1.1. One thin raised stroke.
ⁿSuperscript nU+207F Superscript Latin small letter n. Superscripts and Subscripts block, Unicode 1.1. A raised n, in Unicode since version 1.1.
ʳSuperscript rU+02B3 Modifier letter small r. Spacing Modifier Letters block, Unicode 1.1. The letter that left the least ink on our screen.
ˢSuperscript sU+02E2 Modifier letter small s. Spacing Modifier Letters block, Unicode 1.1. A raised s.
ˣSuperscript xU+02E3 Modifier letter small x. Spacing Modifier Letters block, Unicode 1.1. The letter x at superscript size: put it beside an ordinary x to see the scale of everything here.
How this list was compiled: Built by script on 2026-09-19, not typed by hand. Every character was checked against UnicodeData.txt, version 18.0.0 (https://www.unicode.org/Public/18.0.0/ucd/UnicodeData.txt): the code points and the character names in the notes come from that file, names in sentence case. Block names come from Blocks.txt (https://www.unicode.org/Public/18.0.0/ucd/Blocks.txt) and the Unicode version each character arrived in from DerivedAge.txt (https://www.unicode.org/Public/18.0.0/ucd/DerivedAge.txt), both version 18.0.0. Other names and uses of a character come from the annotations in NamesList.txt, version 18.0.0 (https://www.unicode.org/Public/18.0.0/ucd/NamesList.txt). That is where tsek, the normal tsheg and the non-breaking tsheg, the CNS 11643 heading of the Small Form Variants block, the Mandarin tones, Avestan and Samaritan, midpoint, Georgian comma and ano teleia, the users of the Canadian syllabics dot, mijaket, the hyphenation example, the multiplication dot, black small circle, the uses of the modifier letter prime, the half rings, ypodiastoli, backtick, the smooth and rough breathing marks (annotated on U+0313 and U+0314), the tone marks, aristeri keraia, the preference for U+2019, minutes and feet, the stress marks, overline and APL overbar, Budu, the phonetic uses of the plus sign, minus sign, arrowheads and half triangular colon, the Swedish grave accent, square bullet, the saltillo, the modifier letter apostrophe and the okina come from, and the sentence that Tiny and Miny yield infinitesimal values in combinatorial game theory. Which characters have an emoji form, and the text-style sequences that keep them plain, come from emoji-data.txt (https://www.unicode.org/Public/18.0.0/ucd/emoji/emoji-data.txt) and emoji-variation-sequences.txt (https://www.unicode.org/Public/18.0.0/ucd/emoji/emoji-variation-sequences.txt), version 18.0.0. Which characters are wide comes from EastAsianWidth.txt, version 18.0.0 (https://www.unicode.org/Public/18.0.0/ucd/EastAsianWidth.txt), read on 2026-09-19. Which characters change when text is normalised comes from the compatibility mappings in the sixth field of UnicodeData.txt; normalization form NFKC is defined in Unicode Standard Annex 15 (https://www.unicode.org/reports/tr15/). 30 of the 90 characters here carry such a mapping, and the build checks each one against the NFKC of its own runtime. None changes under NFC: the Greek ano teleia, which does, was left out for that reason. Where each dot of a Braille cell sits (1, 2, 3 and 7 down the left, 4, 5, 6 and 8 down the right) is the numbering of ISO 11548-1 as the Unicode Standard describes it in its section on Braille (https://www.unicode.org/versions/latest/); the drawn positions were checked against it. The order is this site's own and no size is a property of a character: every character was drawn on 2026-09-19 at one size with the system fonts of macOS, the dark pixels were counted, and each group runs from the least ink to the most, checked by eye, with the eight Braille cells kept in dot order. Another font will shuffle neighbours. The same drawing gave where the ink sits (raised, middle, baseline or below) and proved that every character here left less ink than a letter x and that none fell back to an empty box. Left out because they drew as empty boxes there: U+2E33 raised dot, U+2E3C stenographic full stop, U+A78F Latin letter sinological dot, and the Unicode 7.0 shapes named tiny and very small in Geometric Shapes Extended (U+1F78C black tiny square, U+1F797 black tiny diamond, U+1F798 black very small diamond among them). Which characters count as small, the plain tile names and the search words are this site's own choice. Characters that run right to left, combining marks and invisible characters are left out on purpose. All sources were checked on 2026-09-19. Licence: own work