✦ AI-Powered
Python Coding Agent
for Jupyter Notebooks
An AI Agent that understands your Notebook context, generates Python code and hands the control back to you. You decide whether to run, edit or discard it, giving you complete control of your workflow
✓ No signup
✓ 50,000 free tokens
✓ Data privacy
For the full feature set, see the Features section.
Ready, steady,
launch
It's just another pip install!
Getting started with Jupyter Express is a breeze. Simply install it like any other Python package, import and launch it in your notebook and you are ready to go. No extra configuration, no API key hunting and nothing to sign up for
50K tokens to get you started
Start exploring Jupyter Express right away with free tokens, assigned to you automatically on first use - no registration required
Features
01 — Transparent execution
Control your workflow
Describe your task and Jupyter Express writes clean Python code directly into your notebook cell. No automatic execution, no surprises. You decide when to run it, keeping you in full control at every step.
02 — Context injection
Reference live variables without typing a single name
Type @ anywhere in the prompt area to open a live browser of your kernel’s current variables and columns, no copy-pasting long names, no typos.
03 — Prompt history
Reuse past prompts, skip the retyping
Every prompt you’ve sent is saved per notebook and stays intact through kernel restarts and crashes. Scroll back through your history or jump straight to the one you need, one click to clear it all when you’re done.
04 — Agent memory
Pick up exactly where you left off
Jupyter Express keeps a persistent conversation history tied to each notebook. Restart your kernel, close your browser, lose power — the context is still there when you return. Wipe it clean anytime with a single click.
Simple Token-Based Pricing
Pay only for what you use. No subscriptions, no hidden fees. Your tokens never expire.
Frequently Asked Questions
To generate runnable Python code, Jupyter Express first combines your prompt, conversation history, and current notebook context (list of variables/columns and their types) with the system prompt. It then sends it to an AI model, receives the reply, formats it, and pastes it into a cell in your Jupyter Notebook.
We constantly evaluate and optimize the AI models powering our code generation, taking into account factors such as accuracy, context window size, and generation speed to ensure the most accurate and runnable code.
No, Jupyter Express only produces runnable Python code inside your Jupyter Notebook. Unlike autonomous agents that execute code in the background like a black box, Jupyter Express always puts you in control – you decide whether to tweak the generated code or run it as-is. This ensures that no variables, files, or data are accidentally overwritten or deleted.
Yes. Jupyter Express generates code based on the current notebook context (list of variables/columns and their types) and the conversation history. It never sends your actual data to the model unless you include it manually in your prompt.
Absolutely not. Simply install Jupyter Express like any other Python package, then import and launch it within your notebook. Everything else is handled automatically in the background. You only need to create an account on our website when you want to purchase more tokens.
Tokens are calculated based on the amount of text processed in each request, including both the input (your prompt, conversation history, and notebook context) and the generated output (reply from the AI model). After each interaction, your used and remaining token counts are shown at the bottom of the interface.
Since Jupyter Express processes your active variables, data columns, and conversation history with every prompt, you can reduce token usage by deleting unused variables and clearing the agent memory using the dedicated button in the interface. This is particularly effective when working with large datasets that contain irrelevant columns for your current analysis.
When you reach your token limit, you can purchase more tokens through our website. A link will be provided to you in the interface which takes you straight to top-up options.
No, an internet connection is required for Jupyter Express to be able to function as processing happens in the cloud.
Currently, Jupyter Express is designed to work exclusively with classic Jupyter Notebook. Support for other environments such as JupyterLab and VS Code is on our roadmap.
