Market Research: What is a JupyterHub?

Hi all,

I’m preparing for my Build a BinderHub workshop and I’d like to provide a description of what a JupyterHub is and why it’s important for BinderHub. I’ve been looking at this page and I like the graphic.

I wondered if anyone has a favourite, snappy description of what a JupyterHub is and why it’s so powerful they’d like to share?



Thanks for the prompt!

My definition is usually something like: JupyterHub is a tool that lets people run their own interactive computing sessions on shared hardware. It lets you define the resources, environment, and means of access to these environments. You can deploy this tool on any kind of hardware that you like - whether it is on your own machine or in the cloud.


My one sentence summaries of JupyterHub are: “JupyterHub is a multi-user notebook server.” or “JupyterHub is for you if you need to let lots of people run notebooks without having to install things locally.”


I don’t know how snappy this is, but I describe JupyterHub as one solution to the problem:

  • you have (or have access to) computational resources of some sort
  • you would like to help/allow some humans use those computational resources
  • notebooks are a good UI for your use case(s)

Where “computational resources” and “humans” can take many forms - researchers on HPC, anonymous users on the cloud, students on a university server, etc… Custom Authenticators and Spawners allow mapping any collection of humans onto any kind of computational resources.

The same thing in a more tagliney form:

JupyterHub is a way to help your humans use your computers. With notebooks!


Loving all these answers so far, thank you! :sparkles:

1 Like

BinderHub with persistence. :wink:

A Jupyter server for authenticated users who don’t have to configure Jupyter for themselves. It works the same for everybody, no matter what their local operating systems might be.