What is WOFF? Web Open Font Format Explained

How WOFF Works

WOFF stands for Web Open Font Format, and like EOT before it, WOFF is fundamentally a wrapper rather than a new way of drawing letters. Inside a WOFF file is an ordinary OpenType or TrueType font, the same tables of glyph outlines, character maps, and layout rules a desktop font would carry. What WOFF adds is a container built specifically to move that font across the web efficiently and to satisfy the browser vendors' requirements for an open, license-friendly format.

The core of WOFF is compression. Each table from the underlying font is compressed individually using zlib (the same DEFLATE algorithm behind ZIP and gzip), and the file records both the compressed and the original size of every table. When a browser downloads a WOFF file, it reads the directory, decompresses each table, and reassembles the exact original OpenType font in memory. The result is a smaller download that expands back to a byte-identical font, so nothing about the typeface is degraded in transit.

WOFF also reserves space for metadata: an optional block, stored as XML, where a font's creator can record licensing terms, vendor information, and credits. This was a deliberate design choice meant to reassure type foundries. Rather than lock a font to a domain the way EOT tried to, WOFF simply travels with its license text attached, making the terms visible without preventing the file from working anywhere. Browsers ignore the metadata for rendering; it exists for humans and tools.

In use, WOFF is referenced from a CSS @font-face rule with a format("woff") hint, and the browser fetches it just like an image or stylesheet. Because it decompresses to a standard font, everything the underlying OpenType supports, ligatures, kerning, multiple scripts, works exactly as it would on the desktop. The wrapper is invisible to the designer beyond choosing it as the delivery format.

History and Standardization

WOFF was created around 2009 by type and browser people working together, drawing on earlier compressed-font experiments, and was proposed specifically as the open, cross-vendor answer to Microsoft's IE-only EOT. The design goals were explicit: use existing OpenType data unchanged, compress it, attach license metadata, and add no digital-rights lock that would keep the format out of the open web.

The W3C took up the format and published it as a formal Recommendation, the WOFF File Format, giving it the vendor-neutral standing EOT never achieved. With a real standard in hand, every major browser implemented WOFF, and for the first time designers had a single web font format that worked in Firefox, Chrome, Safari, and Internet Explorer alike. That universality is what finally made custom web typography routine rather than a compatibility gamble.

WOFF's success set the stage for its own successor. Once the industry had agreed on the wrapper approach, attention turned to squeezing the files smaller, which produced WOFF 2.0 with much stronger compression. WOFF 1.0 remains widely supported and is often still listed as a fallback in @font-face rules, but for new work the newer version has largely taken over because it delivers the same fonts in noticeably fewer bytes.

WOFF Compared to WOFF2 and Raw Fonts

Against WOFF2, the difference is compression, not capability. WOFF uses general-purpose zlib on each table, while WOFF2 uses Brotli together with a font-specific transform that understands the structure of glyph data, typically producing files roughly a quarter to a third smaller. Both decompress to the identical OpenType font, so the only trade-off is that a handful of very old browsers understood WOFF but not WOFF2, which is why WOFF lingered as a safety-net fallback.

Against a raw .ttf or .otf, WOFF's advantage is simply size: the same font ships smaller because its tables are compressed. Browsers can load raw fonts, and web servers can also gzip them on the fly, but WOFF standardizes the compression into the file itself and adds the license metadata block, which is why it became the expected delivery form rather than serving desktop files directly.

Against the obsolete EOT, WOFF is strictly better in every way that matters: it is an open standard supported by all browsers rather than one, it carries no domain-locking DRM, and it compresses at least as well. WOFF took EOT's sound idea of packaging a compressed font for the web and delivered it without the single-vendor dependence, which is precisely why it, and then WOFF2, replaced EOT entirely.