2016 in review

2017 is well on its way so it’s about time we took a quick look back at 2016! It has been 7 years since MathJax’s first beta was released - incredible! In that time, MathJax has grown rapidly, both in features, performance and reach, from a stop-gap measure to the de-facto standard for math on the web. And we’ve kept our pace up in 2016 - it’s been one of the busiest years yet.


We were able to celebrate several releases this year. MathJax itself saw both a bug fix release as well as a (rather large) feature release. Our releases were made so much richer by fantastic people contributing a growing number of complex third-party extensions, in particular arabic, mhchem, siunitx, and physics.

MathJax-node has seen a feature release alongside two bug-fix releases. We are also working on on a major jump to v1.0.

By far the most important release for us has been the launch of an entirely new set of tools: our MathJax Accessibility Extensions. These tools form an important milestone for MathJax. Instead of having to rely on partial and low-quality support for math across assistive technology, the Accessibility Extensions provide a simple, drop-in tool that provides universal rendering with the same ease of use for authors and readers as MathJax has provided for visual rendering.

We are grateful for the support of the Alfred P. Sloan Foundation which made this work possible. The expansion of MathJax’s scope is already helping us push the envelope of math on the web for all users. Our research going into the underlying semantic analysis as well as the relevant web technologies has been crucial (and of course available on GitHub). We look forward to continuing the necessary R&D this year - stay tuned for news!

Towards v3.0

In late 2015, we laid out our plans for revamping MathJax’s internals. Thanks to the overwhelming support from our sponsors for these plans, we were thrilled to begin the journey towards this major milestone.

Over the past year, we’ve worked steadily on designing v3.0. This would not have been possible without our Technical Committee, a dedicated group of developers providing us with their input and opinions. Of course, those discussions are available publicly with recordings of our meetings being available on YouTube, minutes on the developer mailing list, and code samples and prototypes on GitHub.

2017 is here!

This year, coding on v3.0 has begun in earnest and you can follow the current developments on GitHub.

CDN usage continues to grow

Perhaps the most reliable sign of MathJax’s success has always been represented by our CDN service. We switched CDN and backend providers twice so far which makes it somewhat difficult to compare the lies statistics. However, for the past few years, we’ve been using CloudFlare so we can compare that data reasonably well.

a bar chart without detailed values
MathJax's CDN unique monthly visitors 2014/08 to 2016/12 (as reported by CloudFlare)

On average, we’ve had ~118 million unique visitors to the CDN each month in 2016. We also went over 150 Million monthly users for the first time in October. Overall, the growth of the CDN is ~63% compared to 2015 (a hair more than the previous year, though the numbers are not directly comparable).

Thank you!

As you can see, it was a great year and we look forward to an even better year in 2017. All of our work is only possible because of our sponsors, our donors and you, our amazing community!

Thanks for a great 2016 and an even better 2017!

The MathJax Team.