Deploy JupyterLab (w/ Codex agent)
JupyterLab with an OpenAI Codex agent that edits and runs your notebooks.
JupyterLab
Just deployed
/home/jovyan
Deploy and Host JupyterLab (w/ Codex agent) on Railway
JupyterLab is the notebook editor for Python. This template adds OpenAI Codex through Jupyter AI: it edits your notebooks, runs cells and commands, and reviews each command before it runs.
About Hosting JupyterLab (w/ Codex agent)
Choose a password and paste your OpenAI API key on the deploy form. The first start installs the agent (about 30 seconds). Sign in, click the chat icon on the left, start a new chat and pick Codex. Everything survives redeploys.
Common Use Cases
- Exploring data with Codex
- Writing and fixing notebooks
- Explaining results
Dependencies for JupyterLab (w/ Codex agent) Hosting
- JupyterLab (Jupyter's scipy-notebook image)
- Jupyter AI with the Codex agent
- An OpenAI API key
Deployment Dependencies
Implementation Details
The agent needs about 0.8 GB of volume and 0.7 GB of memory: use Hobby or higher. Usage is billed to your OpenAI account. Back up the volume.
Why Deploy JupyterLab (w/ Codex agent) on Railway?
Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.
By deploying JupyterLab (w/ Codex agent) on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.
Updated 2026-10-29: first release, JupyterLab 4.6.4, Jupyter AI 3.2.0, Codex ACP adapter 2.3.0 (Codex 0.164.0).
Template Content
OPENAI_API_KEY
Your OpenAI API key, from platform.openai.com (API keys). The Codex agent uses it; usage is billed to your OpenAI account.
JUPYTER_PASSWORD
Choose the password you will sign in to JupyterLab with. Use 12 or more characters: anyone with your URL and this password can run code on your server.
