jupyter-nbmodel-client

What it is and what it does

Jupyter NbModel Client is a Python library that connects to a live Jupyter notebook server via WebSocket and allows you to programmatically manipulate notebooks—adding cells, executing code, and retrieving outputs—without directly managing a kernel process yourself. It acts as a remote control for notebooks running on a JupyterLab server, useful for automation, testing, or building tools that need to interact with notebooks as live objects.

The library depends on jupyter-ydoc, nbformat, pycrdt, requests, and websockets to handle notebook synchronization, serialization, HTTP setup, and WebSocket communication. It supports both standalone JupyterLab instances and Datalayer collaborative notebook rooms. Execution requires an active JupyterLab server, a valid authentication token, and an async-capable Python environment.

Use it for:

• Automate notebook cell creation and execution in a running JupyterLab instance for testing or batch processing.
• Build CI/CD pipelines that execute notebook cells and validate outputs without spawning separate kernel processes.
• Create tools or agents that programmatically generate and run Jupyter notebooks on a remote server.
• Integrate notebook execution into larger Python applications that need to interact with live Jupyter environments.
• Test notebook code and visualizations by adding cells, executing them, and capturing results programmatically.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Programmatically interact with live Jupyter notebooks over WebSocket, adding and executing cells, and reading their outputs without running a kernel directly.

Yes, if you need to automate or control Jupyter notebooks from Python code. The package is actively maintained, has low install friction, carries a permissive BSD license, and requires only standard dependencies. It is most useful for automation, testing, and integration scenarios where you control or have access to a JupyterLab server. Not suitable if you need to run notebooks in isolation without a live server.

Install

jupyter-nbmodel-client on PyPI

pip install jupyter-nbmodel-client

uv add jupyter-nbmodel-client

poetry add jupyter-nbmodel-client

Installing jupyter-nbmodel-client

Before you install

Low friction install with a pure-Python wheel. Active maintenance as of 2026-08-14 with recent releases. Five runtime dependencies (jupyter-ydoc, nbformat, pycrdt, requests, websockets) are all established packages.

License in practice

BSD 3-Clause License permits commercial and private use with attribution and liability disclaimer; no restrictions on modification or redistribution.

Quickstart

from jupyter_nbmodel_client import NbModelClient, get_jupyter_notebook_websocket_url

ws_url = get_jupyter_notebook_websocket_url(
    server_url="http://localhost:8888",
    token="MY_TOKEN",
    path="test.ipynb"
)

async with NbModelClient(ws_url) as nbmodel:
    nbmodel.add_code_cell("print('hello world')")

Requires a running JupyterLab server with a known token and notebook path; async context manager requires an async-capable Python environment (e.g., IPython or Jupyter console).

Verify before relying

• Whether the package works with JupyterLab versions other than those shown in examples
• Performance characteristics when managing many concurrent cells or large notebooks
• Compatibility with Jupyter Server versions and authentication schemes beyond token-based auth

Package facts

Evidence: jupyter_nbmodel_client-1.5.1-py3-none-any.whl

Keywords: Jupyter

Tags

More Software Development packages