Written and maintained by CASRAI Editorial Board
Last updated
To use Anaconda for Jupyter Notebook, you start the notebook server with one command, jupyter notebook, from a terminal, or click the Jupyter Notebook tile in Anaconda Navigator. The server prints a local address (by default http://localhost:8888) and opens your default web browser at it. Jupyter’s own documentation also lists the plain pip install notebook route, and this guide shows when each route fits a research workflow.
What it is used for in research
Anaconda’s Navigator documentation describes Jupyter Notebook as a system that combines code, descriptive text, output, images and interactive interfaces into a single notebook file that is edited, viewed and used in a web browser. Researchers use it to keep an analysis, its output and the explanation of it together. For a lab-record view of notebooks, see the CASRAI guide to free and open-source electronic lab notebooks; for the analysis side, see what is data science.
Install or access (official source only)
How to install Jupyter Notebook
The Jupyter project’s install page uses pip, which it calls the recommended installation tool for Python, and gives this command for the classic Notebook:
pip install notebookThe same page says that if you need environment management rather than just installation, you should look into conda, mamba, pipenv or Homebrew. For that reason the two routes below are a real choice, not two spellings of one step.
| Route | What the official pages say |
|---|---|
| pip | pip install notebook, then jupyter notebook (jupyter.org/install). |
| conda or mamba | jupyter.org/install points to conda and mamba for environment management, and recommends the conda-forge channel when JupyterLab is installed that way. For the difference between the two tools, see mamba vs conda. |
| Anaconda Navigator | Anaconda’s Navigator documentation says Jupyter Notebook is used the same way as the other tools on the Home page: click its tile. |
How to install Jupyter Notebook on Mac
On a Mac the same pip install notebook command applies, and the Jupyter install page also lists Homebrew, which it describes as a package manager for macOS and Linux, as another way to install Jupyter. On Mac and Linux you start the server from Terminal; on Windows, from Command Prompt.
Step-by-step task: open and run Jupyter Notebook
How to open Jupyter Notebook from the terminal
- Open a terminal: Terminal on Mac or Linux, Command Prompt on Windows (per the Jupyter documentation).
- Run
jupyter notebook. - Read the lines the server prints. They include the address it is running at, by default
http://localhost:8888. - Your default web browser opens at that address. If it does not, enter the printed address in a browser yourself.
How to launch Jupyter Notebook from Anaconda
If Anaconda is installed, open Anaconda Navigator and click the Jupyter Notebook tile on the Home page. Anaconda’s Navigator documentation describes launching Jupyter Notebook the same way as Spyder, by clicking its tile.
How to run Jupyter Notebook
Running it means keeping the server process alive in the terminal while you work in the browser tab. The documented options are:
jupyter notebook notebook.ipynbopens a specific notebook.jupyter notebook --port 9999starts the server on a port you choose.jupyter notebook --no-browserstarts the server without opening a browser.jupyter notebook --helplists the server options.- To stop, press Control-C in the terminal; the server’s own message says this stops the server and shuts down all kernels, and pressing it twice skips the confirmation.
Output and how to read it
The terminal output is the server’s log. Its startup line states where the notebook is running (by default http://localhost:8888/) and repeats the Control-C instruction.
Troubleshooting
| Symptom | What the documentation supports |
|---|---|
| Port 8888 is already in use | The server searches for the next available port, so read the printed address rather than assuming 8888. You can also choose one with --port. |
| The browser did not open | Copy the address printed in the terminal into a browser. The --no-browser option exists precisely to suppress the automatic open. |
| You need to shut it down | Control-C in the terminal, twice to skip the confirmation; this also shuts down all kernels. |
| Not sure which options exist | Run jupyter notebook --help. |
Free alternatives (table)
These are the other tools on the Jupyter project’s install page, each with its official command.
| Tool | Install | Launch |
|---|---|---|
| Jupyter Notebook (classic) | pip install notebook |
jupyter notebook |
| JupyterLab | pip install jupyterlab |
jupyter lab |
| Voilà | pip install voila |
voila |
For making an analysis environment citable and repeatable, see the dictionary entry on the computational environment, and the guide to citations in Quarto, R Markdown and Pandoc for a related notebook-style authoring workflow.
FAQ
How do I install Jupyter Notebook?
Run pip install notebook, then start it with jupyter notebook. Both commands are on the Jupyter project’s install page.
How do I open Jupyter Notebook from the terminal?
Type jupyter notebook in Terminal (Mac or Linux) or Command Prompt (Windows). The server opens your default browser at its local address.
How do I launch Jupyter Notebook from Anaconda?
Click the Jupyter Notebook tile on the Anaconda Navigator Home page, or run jupyter notebook from a terminal.
Which port does Jupyter Notebook use?
8888 by default. If it is in use, the server searches for the next available port; --port sets one manually.
How do I stop Jupyter Notebook?
Press Control-C in the terminal that runs the server (twice to skip confirmation). This shuts down all kernels.
Sources (official docs)
- Project Jupyter, Install, checked 2026-10-29.
- Jupyter documentation, Running the Notebook, checked 2026-10-29.
- Anaconda documentation, Anaconda Navigator, checked 2026-10-29.
