WhatFontIs-Bench - A Synthetic Benchmark for Font Family Identification

A synthetic test set for font family identification: a single word, set in a known font, printed or painted on real surfaces and in real scenes, with the exact font, the text and the position of every letter recorded for each image.

Made by WhatFontIs, the font finder that identifies fonts from images, to measure how well a tool can find the font in a real-looking photo.

• Official page: whatfontis.com/WhatFontIs-Bench.html
• GitHub (raw zips, COCO annotations, tools): whatfontis/WhatFontIs-Bench

Paper: A. Cuibari, WhatFontIs-Bench: A Synthetic Benchmark for Font Family Identification in Real-Looking Images, 2026. doi:10.5281/zenodo.22876579

Quick start

from datasets import load_dataset

ds = load_dataset("whatfontis/WhatFontIs-Bench", split="test")
ex = ds[0]
ex["image"]        # PIL image
ex["font"], ex["family"], ex["text"], ex["difficulty"]

v1.0 contents

Fonts	600: 200 sans-serif, 200 serif, 100 slab serif, 100 monospaced
Font sources	215 Adobe Fonts, 191 Creative Fabrica, 183 Google Fonts, 11 DaFont (commercial-use license)
Images	11,995 (20 per font; Colt Soft Regular has 15)
Text	one word per image, 7-12 letters, taken from COCO-Text v2
Size	long side 442-1998 px, short side at least 301 px

Each font has 20 images:

Type	Easy	Medium	Hard	Per font
texture: the word printed or painted on a close-up photo of a real surface (wood, plaster, concrete, brick, metal, fabric, leather, paper, cardboard)	3	2	5	10
scene: the word painted on a wall inside a real photo (rooms, shops, facades, workshops)	1	2	3	6
object: the word on a printed object (label, box, book cover, poster, menu, shop sign, packaging, business card) over a real photo	1	1	2	4

Difficulty levels:

• easy: capital height 160-220 px, little blur and noise, JPEG quality 88-95, almost even light
• medium: capital height 130-180 px, blur 0.3-0.8, noise 1.5-3.5, JPEG 78-90, uneven light
• hard: capital height 100-140 px, blur 0.7-1.3, noise 3-6, JPEG 65-80, strong uneven light, and often a cast shadow over part of the word, a glare spot or light print wear

The images are frontal, with no slant or perspective, so an upright font never looks italic. Text contrast is at least 70 grey levels (50 on the hard level). No letter is clipped. Fonts that only have capitals (107) are written in capitals.

Columns

Column	Meaning
image	the JPEG image
id, file_name	image id and original path (scenes/NNNNN.jpg)
text	the word in the image
font, family, category, font_source, font_id	the font: full title, family, class (Sans-Serif / Serif / Slab Serif / Monospaced), where it comes from, WhatFontIs id
type, difficulty, surface	texture / scene / object, easy / medium / hard, surface kind (texture images)
caps_only, effects, cap_height_px, width, height	capitals-only font flag, hard-level effects, capital height and image size in px
ink_quad	four corners of the word (TL, TR, BR, BL)
char_quads	four corners of every letter, in reading order
crop_box	[x0, y0, x1, y1] of the text crop

The full per-image record (camera transform, blur, noise, JPEG quality, ink colour, background source and license) is in v1/labels.jsonl; the same set in COCO format is in v1/annotations.json. Also included: v1/fonts.json (the 600 fonts), v1/backgrounds.json (the 153 background photos with author and license), v1/specimens/ (sheets showing all fonts), words.txt and tools/make_crops.py. Font files are not included.

Evaluation and baseline: WhatFontIs API

A result counts as correct when it is the right font family, in any weight or from any source: an image in Roboto Bold answered with Roboto Regular is correct, answered with Arial Bold it is wrong.

Results of the WhatFontIs API (api2) on all 11,995 images, September 2026. The API searched its whole catalogue of over 1.2 million fonts, not only the 600 fonts of this set.

	Images	Correct at 1	In top 5	In top 20
All images	11,995	83.7 %	93.3 %	96.5 %
texture	5,997	83.6 %	93.3 %	96.3 %
scene	3,599	82.4 %	92.3 %	96.0 %
object	2,399	85.9 %	95.0 %	97.5 %
easy	2,995	85.2 %	93.9 %	97.4 %
medium	3,000	83.3 %	92.8 %	96.0 %
hard	6,000	83.1 %	93.3 %	96.3 %
sans-serif	4,000	75.7 %	88.7 %	94.2 %
serif	4,000	84.0 %	95.1 %	97.7 %
slab serif	1,995	95.0 %	98.5 %	99.0 %
monospaced	2,000	87.8 %	94.0 %	96.1 %

Known gaps

• Colt Soft Regular has 15 images instead of 20 (too wide for a 7-letter word at the easy size within 2000 px).
• The scene images reuse a limited number of real scenes, each with a different word, font, position and light.
• Some fonts in the set are close look-alikes of other fonts that exist under different names.

Licenses and credits

• Dataset (images and annotations): CC BY 4.0. Please credit WhatFontIs-Bench (WhatFontIs) with a link to whatfontis.com/WhatFontIs-Bench.html.
• Backgrounds: all CC0 1.0. Surface photos from Poly Haven and ambientCG; scene photos from Wikimedia Commons, found through Openverse. Per-file details are in v1/backgrounds.json.
• Words: from COCO-Text v2 (CC BY 4.0), Veit et al., COCO-Text: Dataset and Benchmark for Text Detection and Recognition in Natural Images, 2016.
• Fonts: the typefaces belong to their designers and foundries. The images show text rendered in them; the font files are not part of this set.

Citation

@misc{cuibari2026whatfontisbench,
  title  = {WhatFontIs-Bench: A Synthetic Benchmark for Font Family Identification in Real-Looking Images},
  author = {Cuibari, Alexandru},
  year   = {2026},
  doi    = {10.5281/zenodo.22876579},
  url    = {https://doi.org/10.5281/zenodo.22876579}
}