Marvin XR

TSS#125: Top AR Frameworks to Use for WebAR development

If you're planning to dive in WebAR in the near future, whether for learning & development purpose or for starting a business with, choosing the right technology can make or break your success. Over the last few years, few frameworks (WebAR development libraries) established themselves as the frontier in the industry. If this is your first AR encounter, then the following list can definitely help you. However, if you're a developer, then you can try to build your own custom framework too.

Web based AR apps Framework List

AR.js

AR.js is a free and open-source AR development framework for web-based applications.

● AR.js is entirely JavaScript-based and needs no installation

● The framework is super-fast and works well on mobile devices as well

● The framework is open source and completely free of cost

● AR.js works on any phone with webGL and webRTC (JavaScript APIs)

AR.js offers the following types of augmented reality development:

● Image tracking, which lets devices recognize and track 2D images and superimpose content onto them

● Marker tracking, which triggers AR experiences by scanning a marker (like a QR code) via a mobile device

● Location-based AR for AR experiences that use a user’s position in the real world to show AR content

AR.js has a massive repository of documentation and a thriving online community, which means you get all the help you need to build AR apps that floor your audiences.

Three.js

Three.js is another most popular kid in the WebAR school. The vast community around it and so many applications being built with it proves the same. This is also a lightweight JS library which is quite easy to use. Some major benefits of using three.js are:

  • Easy to learn

  • Plenty of examples

  • Large community, lots of third-party plugins

  • Good documentation

  • Great performance

  • PBR rendering

  • Support for most of the popular model formats

  • Has own editor

However, building a complex webAR app with three.js can be pretty complex in terms of writing codes for it. If you are not super savvy in JS or have not learnt three.js deeply, diving deep could be a red flag for your startup.

model viewer

If you want to build a SLAM based WebAR experience, model viewer can be your friend. Using open source model viewer js library is a bliss for them who want to build WebAR while writing not more than 10 lines of code in the first place. It works easily with iOS ARKit and Android ARCore. The long list of device compatibility is a bliss.

Cons:

model viewer only offers SLAM WebAR and not the other types.

Mind AR

This is a relatively new JS framework for WebAR. The prominent features for Mind AR are image tracking and face tracking WebAR. It has a good community support to begin with. The Github codebase can be used to develop your first WebAR app.

Bonus List for AR app development framework

If you are looking for building a native AR apps, then the following framework could be of great help to you:

ARKit – apps for iOS

ARKit is Apple’s AR development framework for iOS devices. It combines motion tracking, world tracking, scene understanding, and display conveniences for cohesive AR experiences on iOS devices.

ARCore – Best for Android apps

ARCore (a.k.a. Google Play Services for AR) is Google’s development platform for augmented reality apps. It’s built on the same capabilities:

1. Motion tracking

2. Environmental understanding

3. Light estimation

Meta Spark Studio – Best for Meta platforms

Meta’s Spark Studio is an AR development platform for Facebook and Instagram. Unfortunately Meta has announced a planned shutdown of this platform in the beginning of 2025.

Camera Kit – Best for Snap platforms

Camera Kit is Snap’s AR platform that lets businesses build AR experiences for Snapchat. AR content built on Camera Kit can be shared with Snapchat, Snap Spectacles, as well as web and mobile apps.

Vuforia – Best for industrial enterprise use cases

Vuforia is an AR development suite built on computer vision technology that is able to recognize and track 2D images and 3D objects in real-time.