---
name: Sans-serif
slug: sans-serif
category: typography
status: published
created: 2026-08-21T00:00:00.000Z
modified: 2026-08-21T00:00:00.000Z
definition: A typeface drawn without finishing strokes on the ends of its stems,
  which is the default register of most interface type.
aliases:
  - name: sans
  - name: grotesque
    source: community
  - name: gothic
    source: community
tags:
  - fonts
relations:
  contrastWith:
    - serif
  variantOf: []
  partOf: []
  seeAlso:
    - superfamily
implementations: []
sources:
  - title: A Glossary of Typographic Terms (Monotype)
    url: https://www.monotype.com/resources/typographic-terms
demo: inline
exhibit: false
useWhen: naming type without finishing strokes
---

The name is a subtraction: French *sans*, without, plus the stroke that is not
there. That is the whole definition, and it is why the category is so wide.
Nineteenth century foundries cut these faces for posters and headlines, not for
books, and gave them names that record how strange they looked to eyes trained on
serifs. English foundries called them grotesque, from the Italian *grottesco*,
meaning cave-like or bizarre. American foundries called them gothic, for no
better reason than that gothic already meant "old and severe". Both words are
still in live use: Franklin Gothic and Akzidenz-Grotesk are the same class of
thing, named twice.

Three sub-genres do most of the work in practice. A grotesque or neo-grotesque
(Helvetica, Arial, Inter) closes its apertures and holds every letter to nearly
the same width, which is why a paragraph of it reads as an even grey and a long
run of it reads as anonymous. A humanist sans (Frutiger, Open Sans, Lato) keeps
the skeleton of a written letter: open apertures, varying widths, a two-storey
lowercase g in some cases, and it stays legible at small sizes because its
letters stay distinguishable from each other. A geometric sans (Futura, Poppins)
draws bowls as near-circles and stems as near-lines, which is beautiful on a
title and tiring across three paragraphs.

Interfaces default to sans for reasons that are mostly historical and partly
practical. Every platform's system face is a sans (San Francisco, Segoe UI,
Roboto), so `system-ui` resolves to one and inherits the hinting and the
optical sizes the platform ships. Small text at low resolution punished hairline
detail, and interface text is mostly small. None of that makes sans inherently
more readable, and the reflex is worth noticing: sans is the safe choice, not the
correct one.

The common pairing convention is one serif and one sans, with the contrast used
to separate registers rather than for decoration: a serif headline over sans body
copy, or sans labels and controls around serif prose. Many families now ship both
halves under one name (Source Serif with Source Sans, IBM Plex Serif with IBM
Plex Sans), which is the low-risk way to get a pairing whose proportions already
agree.