Pairing font is a constant challenge for designers and typographers. Putting the wrong fonts together can create a discordant look that ruins a design, while letters that share similar themes or measures can bring a project together. To help creatives match their fonts successfully, Fontjoy has turned to deep learning to make the whole process quicker than ever.
Created by designer and engineer Jack Qiao, Fontjoy builds on a similar font mapping tool released earlier this year. By extracting the feature vectors from images of nearly 2,000 fonts, Qiao was able to create a formula that can systematically sift through styles and find fonts that share key characteristics.
On Fontjoy you can play around with font pairings by selecting styles from the options in the sidebar. Simply choose how similar or contrasting you want the fonts to be by adjusting a scale at the top of the site, then click generate to instantly find a match.
The controls take a bit of getting used to, but once you get the hang of Fontjoy you'll be able to lock you favourite fonts and experiment with pairings in no time.
For experienced designers Fontjoy might prove to be little more than a starting point for your own experimentation, but for beginners the site is a useful tool in understanding how font styles can work together or against each other. And with Qiao explaining how he created the mapping algorithm on Github, there's even some insights for those who want to start getting into machine learning.