Scripts beyond Latin: who covers what
Of the 1,899 measured families, 1,177 (62%) cover at least one script subset beyond Latin. Vietnamese is the most common addition with 532 families, then Cyrillic (306) and Greek (128); 670 families stop at Latin and Latin Extended.
What a subset is
Google Fonts splits each family into subsets by script so that a page loads only the characters it needs. The subsets a family supports are listed in its metadata. A subset says the characters are there; it does not say how many languages written in that script are fully served, and Latin Extended in particular varies from one family to the next. The library uses 180 subsets in all.
| Subset | Families | Share of the library | Median glyphs |
|---|---|---|---|
| Vietnamese | 532 | 28% | 1,000.5 |
| Cyrillic | 306 | 16% | 1,321.0 |
| Cyrillic Extended | 243 | 13% | 1,123 |
| Greek | 128 | 7% | 1,942.0 |
| Greek Extended | 76 | 4% | 2,304.0 |
| Math | 73 | 4% | 713 |
| Symbols | 69 | 4% | 687 |
| Japanese | 69 | 4% | 8,073 |
| Devanagari | 65 | 3% | 1,060 |
| Hebrew | 62 | 3% | 774.0 |
| Arabic | 54 | 3% | 1,228.5 |
| Korean | 47 | 2% | 12,103 |
| Thai | 34 | 2% | 773.0 |
| Telugu | 27 | 1% | 1,234 |
Cyrillic, Greek and Vietnamese
These three extend Latin-based families most often because they share many shapes with Latin. Of the 306 families with Cyrillic, 139 are sans serifs and only 15 are handwriting. Greek appears in 128 families. Vietnamese, which needs Latin letters with stacked diacritics, is in 532; it is the one extension that handwriting families carry often (100 of them).
Large character sets
Scripts change the size of a font more than anything else. The median family with basic Latin has 608 glyphs in its regular file; with Cyrillic, 1,321.0; with Greek, 1,942.0. Japanese families have a median of 8,073 glyphs and Korean families 12,103. Of the 69 families with a Japanese subset, the earliest was added in 2015.
Basic and extended subsets
Cyrillic and Greek each come in a basic and an extended subset. Of the 306 families with basic Cyrillic, 215 also list Cyrillic Extended, which adds letters for further languages written in Cyrillic; 64 of the 128 Greek families list Greek Extended, which carries the polytonic accented letters. Vietnamese reaches 45% of monospaced families and at least 22% of every class.
Arabic, Hebrew and the scripts of South and Southeast Asia
Scripts that are not related to Latin are covered by far fewer families. Arabic appears in 54, Hebrew in 62 and Devanagari in 65. These families almost always keep a Latin subset as well, so the same file can set a bilingual page. Display faces are the largest group for Hebrew (27 families), while Arabic and Devanagari lean towards sans serifs. The table gives the earliest year in which a family with each subset was added; a family that gained the subset later counts from its own date.
| Subset | Families | Largest class | Earliest family added | Variable |
|---|---|---|---|---|
| Arabic | 54 | Sans-serif (26) | 2010 | 29 |
| Hebrew | 62 | Display (27) | 2010 | 15 |
| Devanagari | 65 | Sans-serif (29) | 2012 | 13 |
| Thai | 34 | Sans-serif (21) | 2013 | 6 |
| Bengali | 11 | Sans-serif (5) | 2015 | 6 |
| Tamil | 15 | Sans-serif (8) | 2015 | 7 |
| Telugu | 27 | Sans-serif (16) | 2014 | 5 |
| Khmer | 24 | Display (11) | 2010 | 6 |
Chinese, Japanese and Korean
A Japanese subset is listed by 69 families, 47 a Korean one and 29 at least one Chinese subset. Their files are the largest in the library: Cactus Classical Serif holds 58,570 glyphs, and the smallest Japanese file still holds 360. Of these families, 22 (15%) are variable fonts, against 29% of the whole library, and 7 of the Japanese families are handwriting designs.
Scripts by class
Coverage beyond the Latin-based subsets differs by class. Sans-serif families reach other scripts most often (65%), handwriting faces least (18%). The table counts a family once if it has any subset other than Latin, Latin Extended and Vietnamese.
| Class | Families | With another script | Share | Median subsets |
|---|---|---|---|---|
| Sans-serif | 732 | 477 | 65% | 3.0 |
| Serif | 354 | 200 | 56% | 3.0 |
| Display | 472 | 163 | 35% | 2.0 |
| Handwriting | 297 | 54 | 18% | 2 |
| Monospace | 44 | 24 | 55% | 3.0 |
How coverage is spread
Of the 180 subsets in use, 18 appear in 20 families or more, 45 in between two and nineteen, and 117 in a single family; 114 of those single-family subsets belong to a family of the Noto project. A further 52 families list no subset at all in their metadata; they are counted in every other measure but not here.
The broadest coverage
Google Sans lists 27 subsets, more than any other family. Next are Iosevka Charon, Iosevka Charon Mono, Google Sans Code and Cascadia Code. Many single-script families from the Noto project cover one script each and are grouped on the Noto Sans page where it is listed.
Rare scripts
Subsets carried by a single family include Ahom, Anatolian Hieroglyphs, Avestan, Bamum, Bassa Vah, Batak, Beria Erfe, Bhaiksuki. Scripts with 20 or more families have their own page; the full list of subsets shows the families behind every smaller one.
Using the list
When a project needs one family for several scripts, filter the type finder by subset first and then by class or weight range. The result counts how many families meet all conditions at once, from the same snapshot of 28 October 2026.
Sources
Family metadata (METADATA.pb) and font files from the google/fonts repository as of 28 October 2026; all measurements computed by this site as described in How we measure.