Connecting Jupyter Notebook to kernel running on a server

Support Help
1

Hi my work set up is as follows:

Windows machine is how I access all the GUIs for my code, but I am using putty to ssh into a server where all my code lives. I have my server mounted to my windows computer via a network drive. I edit a combination of jupyter notebooks and just regular code. For the code part, I have cursor set up and working for that use case. However, I am still trying to figure out the notebooks. My server is quite a bit more powerful than my computer and has a special python environment installed on it, so I need to use the server to run notebook cells. How do I do this from Cursor? Normally I run a jupyter notebook and then access it from chrome.

2

Hi Amit,

I was able to figure this out using Claude Haiku on t3.chat.

Here are the instructions I followed:

Step 1: Start the Jupyter server on your remote machine

  • Run jupyter notebook, jupyter lab, or open a notebook through Anaconda

  • Copy the server URL and token from the terminal output (looks like http://localhost:8888?token=abc123…)

  • If u can’t find it, dump the output into a LLM and itll find it for you

Step 2: Connect Cursor to the remote server

  • Open a .ipynb file in Cursor

  • Click the kernel selector (top of the notebook)

  • Select “Existing Jupyter Server”

  • Paste the server URL from Step 1

  • If it’s asking for a password, check your terminal log for the entirety of the token.

Step 3: Run your notebooks

  • Cursor will now execute cells on your remote server using its Python environment

  • Your files are already accessible via the network drive

Good luck!