Artificial Intelligence (AI) and Large Language Models (LLMs) play an increasingly important role in the industry, and within the Jupyter community. This ecosystem is also relatively young and evolving rapidly. The goal of this page is to point you in the right direction to learn more about various initiatives within the Jupyter ecosystem related to AI.
Technical efforts related to AI in Jupyter subprojects
Jupyter AI
Jupyter AI is an extension that connects AI agents to computational notebooks in JupyterLab. It is currently incubating as a part of the Jupyter-Frontends subproject.
Jupyter AI Lite
Jupyter AI Lite provides AI code completions and chat for JupyterLab, Notebook 7 and JupyterLite, with a persona named Jupyternaut running in the browser. It is maintained alongside Jupyter AI in the JupyterLab organization.
Jupyter AI Contrib
jupyter-ai-contrib is a community organization hosting extensions, integrations, and experiments that compose or build on Jupyter AI. It is a combination of earlier-stage projects that undergo more rapid development and prototyping, and community-maintained projects.
Where to discuss AI in the Jupyter community?
For asynchronous chat, the JupyterLab channel in Zulip is your best bet. Synchronous conversations and demos happen during the weekly Jupyter AI meeting (see calendar) and community workshops. Because the Jupyter AI project is incubating within the Jupyter Frontends subproject, this channel is the best place to have developer-related conversations about AI.
Frequently Asked Questions
What is the difference between Jupyter AI and Jupyter AI Lite?
Both projects integrate AI features into JupyterLab and Notebook v7+ interfaces. Jupyter AI Lite does not depend on a server component, which makes it possible to use it with JupyterLite, the in-browser Jupyter distribution. Jupyter AI enables agentic workflows that can continue to run on the server even when the browser is closed or disconnected. Server dependency distinction aside, the set of features overlaps, with both packages implementing agentic workflows and tool calling, and Jupyter AI Lite reuses the chat interface and persona manager of Jupyter AI, so the user experience is the same in the browser and on a server. Notably, Jupyter AI Lite also implements AI code completion, and Jupyter AI implements multi-agent workflow.
Does Jupyter have an AI contribution policy?
Not yet - Jupyter is primarily organized into semi-independent subprojects, each of which has their own practices and policies for contributing. There is no Jupyter-wide policy regarding AI-assisted contributions. Many sub-projects are currently in the process of defining the policies that work for their needs and practices.
Is there a single “AI subproject” in Jupyter?
No, Jupyter has a number of initiatives related to AI as part of subprojects (linked above), but there is no one AI subproject that encompasses everything. We’re encouraging subprojects to experiment and develop their own AI-related work, and to integrate and connect with other pieces of the ecosystem.