#3 · Primary category: Web Performance & Predictive Loading
guess
🔮 Libraries & tools for enabling Machine Learning driven user-experiences on the web
Project last updated:08/27/26
GitHub Stars
7.1K
Forks
198
Contributors
27
License
MIT
Why we included this project
Guess.js targets a narrow, concrete problem: the wait a user feels moving between pages of a JavaScript app. Instead of prefetching everything, it reads analytics data, usually from Google Analytics, to learn which page a visitor is most likely to open next, and pulls in that resource ahead of time. The GuessPlugin handles most of the setup at build time for webpack users, while a separate workflow covers static sites that skip webpack. Teams that care about navigation speed and want data-driven prefetching instead of a blanket preload-everything approach will find what they need here. The project is still alpha and only makes good calls once it has real traffic data to learn from, so plan on seeding analytics and testing before you trust it in production.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
lazysizes
High performance and SEO friendly lazy loader for images (responsive and normal), iframes and more, that detects any visibility changes triggered through user interaction, CSS or JavaScript without configuration.
vanilla-lazyload
A lightweight vanilla JavaScript lazy loading library that defers images, videos, iframes, and scripts until they enter the viewport, using IntersectionObserver and supporting responsive images.