Find files, editable templates and browser test targets by what you need to make or test. The directory below is cut by format; the two collections under it cut the same library by subject and by workflow.
The latin basic subset in the container a site would serve, so a size comparison against the full WOFF2 measures the subset rather than the compressor. Original typeface, free to use.
The same CFF (Type 2) face in a WOFF2 container, 45 percent of the raw sfnt size. Compare it against the other three containers in this group to see how much of a web font's weight is the format and how much is the design. Original typeface, free to use.
The COLR v0 colour font as WOFF2, for testing that a web-font pipeline keeps COLR and CPAL rather than stripping them as unknown tables. Original typeface, free to use.
The COLR v1 gradient font as WOFF2: the shape a site would actually serve. For testing that gradient paint graphs survive web-font compression intact. Original typeface, free to use.
The all-features font as WOFF2, for testing that a web-font build step preserves GSUB, GPOS and GDEF instead of subsetting the layout tables away by default. Original typeface, free to use.
The full face as WOFF2: the honest baseline for a subsetting benchmark, since comparing a subsetted WOFF2 against an unsubsetted TTF measures Brotli rather than the subsetter. Original typeface, free to use.
The Novus Sans Sample demonstration typeface as WOFF2 (Brotli-compressed web font): one member of a TTF/OTF/WOFF/WOFF2 conversion set built from scratch with fontTools. Original and free to use; a fixture for testing font loading, subsetting, and conversion.
The same TrueType (glyf) face in a WOFF2 container, 33 percent of the raw sfnt size. Compare it against the other three containers in this group to see how much of a web font's weight is the format and how much is the design. Original typeface, free to use.
The four-axis variable font as WOFF2. For testing web font loaders and CSS font-variation-settings against a face with more axes than a typical release. Original typeface, free to use.
The wght variable font wrapped as WOFF2: the same fvar/gvar payload in the web container, for testing that a compressor or a CDN pipeline preserves variation data. Original typeface, free to use.
The two-axis variable font as WOFF2: the container a real site would ship. For testing that variation tables survive Brotli packing and that a two-axis slider still works after decompression. Original typeface, free to use.
A WOFF2 file carrying the optional extended-metadata block from the WOFF specification: vendor, credits, description and licence as XML alongside the font, outside the sfnt tables. Most converters drop it silently; this is the file that proves whether yours does. Original typeface, free to use.
An adjustment request built around what somebody needs rather than why, so it collects no health information at all. Useful as an accessible-form reference and as a counter-example to intake forms that ask for a diagnosis.
The same 40 world cities as a GeoJSON FeatureCollection of Points (longitude, latitude order per RFC 7946), each with name, country, and population properties. The mapping twin of the CSV, for testing GeoJSON parsers and map renderers.
A curated world-cities dataset: 40 major cities with country, latitude, longitude, and population. A realistic geospatial fixture for testing map plots, geocoding, and CSV→GeoJSON conversion. GeoJSON twin included.
The same plate encoded once at JPEG quality 25. Blocking and ringing are bounded by that single pass, so a restorer can be scored against the PNG reference in this group rather than against an opinion Plate pr-watch.
The same plate encoded once at JPEG quality 50. Blocking and ringing are bounded by that single pass, so a restorer can be scored against the PNG reference in this group rather than against an opinion Plate pr-watch.
The same plate encoded once at JPEG quality 75. Blocking and ringing are bounded by that single pass, so a restorer can be scored against the PNG reference in this group rather than against an opinion Plate pr-watch.
The same plate encoded once at JPEG quality 90. Blocking and ringing are bounded by that single pass, so a restorer can be scored against the PNG reference in this group rather than against an opinion Plate pr-watch.
A synthetic photographic plate of a single product on a plain studio ground, 1024x1024, lossless PNG. This is the reference every lossy variant in this group is derived from Plate pr-watch.
