Google Colab (which is short for Colaboratory) is a hosted Jupyter notebook. It is a great option for data scientists and python library authors because:

• It is free to use.
• It is easy to open notebooks directly from GitHub.

If you’re browsing Github for notebooks, you’ll find a lot of links to open them on Google Colab for this reason.

Because it’s based on Jupyter, you can use the vast array of scientific computing libraries available on Python. However, just like Jupyter, it’s got a bit of a steep learning curve. It works best for technical people who are comfortable with Git.

It has rich integration with Google Drive. You can load files directly from Google Drive which is especially convenient if you have datasets in Google Sheets. If you’re not opening notebooks from GitHub, you can conveniently open notebooks from and save to Google Drive, meaning notebooks live alongside your other Google Drive files.

Despite “colab” being in its name, it’s not the best option for teams. It lacks realtime collaboration and versioning, which means you’ll have a much harder time working with other people.