Run on Gradient

Introducing a new way to launch a GPU-backed notebook from any .ipynb file in any public GitHub repository

Read the docs

Join us on our mission to make Jupyter notebooks across the web easy to fork, easy to run, and easy to scale into meaningful applications.

For Builders

We'll help you grow your community by providing simple ML infrastructure for others to run your notebooks.

And Explorers

Together we can make it easier to share your ML projects and help others get a running start.

With No Fuss

Once you build the link, your Run on Gradient template is guaranteed to work for as long as you maintain its repo.

We've helped hundreds of thousands of developers launch machine learning and deep learning projects with free CPUs and GPUs



  • CPU only

  • 2GB RAM

  • 2 vCPU




  • NVIDIA M4000 GPU

  • 30GB RAM

  • 8 vCPU




  • NVIDIA P5000 GPU

  • 30 GB RAM

  • 8 vCPU


Gradient Notebooks support every popular library and framework out there

To get started, just build yourself a link:

Start with the base Paperspace console link for GitHub

Identify an .ipynb file in a public GitHub repository

Combine the two

And then inject the link into a button one of two ways:




<a href="">
 <img src="" alt="Run on Gradient" />

Or try the link generator!

Running a notebook from a Run on Gradient badge in GitHub