Skip to main content

​
General

Jupyter Notebooks provide a web-based interface for creating and sharing computational documents. You can seamlessly mix executable code, documentation, and instructions in one portable document. Notebooks are not only a great portable learning tool, but also a highly capable vehicle for prototyping and producing data science work.Anaconda Notebooks lets you skip setup and installation and get straight to learning and writing code.
You can access and use Anaconda Notebooks from any modern web browser and anywhere you have an internet connection.After you have logged into your account on Anaconda.com, navigate to nb.anaconda.com or click on the “Notebooks” tile on Anaconda.com.
Anaconda Notebooks provides the following features, hosted on our trusted and secure cloud platform, so you can access them anywhere, on any device.
These features are limited based on your subscription plan. See our pricing page for further details.
  • A curated JupyterLab notebook interface
  • Fast, backed-up SSD storage
  • CPU seconds, refreshed daily
  • Anaconda Assistant, your AI-powered pair programmer
  • Application publishing
  • Conda environments with the most popular python packages
  • Ability to create and upload your own custom environments
  • Example notebooks to develop your coding skills
Anaconda Notebooks is a hosted JupyterLab service that enables you to run JupyterLab notebooks reliably online. Your dedicated JupyterLab instance comes pre-configured with persistent cloud storage, hundreds of data science packages, and a managed infrastructure.To launch an instance of classic Jupyter Notebooks, click Help in the menu bar, then select Launch Classic Notebook.
You can get community support on the Anaconda community forums. If you’re in need of further technical assistance, please file a support ticket.
All packages available from the Anaconda installer are preloaded and ready to use through Anaconda Notebooks. The latest release of Anaconda Distribution is always the default environment within Anaconda Notebooks. As new installers are released, new environments become available. For more information about managing environments and notebook runtimes, see Runtimes.Anaconda Toolbox also comes with several pre-configured environments, which include popular data science packages tailored for different analytical needs.To see a list of all packages in an environment, run the following command from a terminal in Anaconda Notebooks:
conda list
Yes! Click Share at the top of your notebook to produce a shareable link or embeddable HTML for your notebook. See Sharing Anaconda Notebooks for more information.
In the Anaconda Notebooks JupyterLab interface, click Upload files in the File Browser to browse for a local .ipynb file. Then, click Open. The notebook will appear in the left-hand menu.
You can also drag and drop a notebook from a folder on your system to the file browser to upload it.
Like most IDEs or editors, JupyterLab has the standard “Save” and “Save As…” functions that will save a notebook in your directory on our platform. You can also download a notebook file from the File menu to save it locally.
Packages available from Anaconda Notebooks are a subset of packages available from the public repo.anaconda.com . Installing packages from a curated premium repository via tokenized access is not currently supported.
You can create your own conda environments to build a custom runtime using any packages that conda can install from repo.anaconda.com. This can be achieved by following the steps in Runtimes.
Anaconda Notebooks alone does not provide commercial compliance to its users. Customers accessing Anaconda Notebooks with Pro (legacy), Business, or Custom subscription plans are permitted to use all Anaconda products for commercial use. For more details, see our Terms of Service.
Registered customers who are part of organizations on Anaconda.com can independently access Anaconda Notebooks. Access to Anaconda Notebooks is granted upon member role designation and registration.
Yes. You can prevent your organization members from accessing Anaconda Notebooks from your organization’s Org Profile page. For more information, see Organization settings.
If you are a customer but have not yet registered your organization on Anaconda Cloud, please refer to Organization management on how to set up your organization and invite members.
The ipykernel is available from the main channel.
Publishing on Anaconda Notebooks provides you with a server-hosted app, while PyScript.com offers you a browser-hosted app.Panel supports both server and browser operation; however, browser-side operations require copying all the data down to the browser, which is impractical when working with large datasets. Additionally, some applications can’t be run browser-side because some libraries use operations that are not available in WASM (for example, libraries like numba, dask, or pytorch). In other words, it’s a matter of running on the server or running locally in your browser.
Data catalogs are sample datasets that you can use to familiarize yourself with running and analyzing datasets in a notebook. Access Anaconda’s catalogs of pre-populated datasets by selecting Explore Catalogs in the Launcher.
To activate the debugger, click the bug icon to the right of your active notebook. Learn more about the debugger from JupyterLab’s Debugger topic.

​
Toolbox environments

No, quick start environments include a list of packages to install, but rely on the conda solver to select the most up-to-date versions that are compatible with the other packages and their dependencies. After installation, you can verify installed package versions by running !conda list in a notebook cell in the active environment. You can also view package versions on the Environments page in Navigator.
While security information isn’t directly accessible, installation security is managed through channel configurations. When installing a quick start environment, conda uses the channels list available in your .condarc file. For customers on the Business plan, this means that you are only able to install quick start environments if your organization’s policies allow it.
All environments require either the Anaconda defaults channel or conda-forge to be included in your .condarc file. You can update your channels in Navigator or with conda via the CLI.
The python-starter environment, which includes the fewest packages, is the quickest to install and should complete in fewer than five minutes. The anaconda-ai-ml environment, on the other hand, could take up to 15 minutes to install, due to the large size of some of the packages and complexity of the package dependency resolution process. Exact installation times vary depending on your internet download speeds and hardware specifications.
Yes! You can install one or all of these environments, provided your computer has the available disk space.
To stop or interrupt a quick start environment during installation, completely close your Jupyter instance, then reopen it via Navigator or the command line.To close a Jupyter instance:
  • In the JupyterLab interface, go to File > Shut Down.
  • Alternatively, return to the shell application where you launched jupyter-lab and press Ctrl+C.
You can continue working in the same notebook, but you’ll need to select an existing runtime. This can be done through the runtime selector, the Launcher, or by using the terminal. Previously executed cells must be re-run in the new runtime.

​
Anaconda Assistant

Some options for working with DataFrames are hidden until you run cells in your notebook that generate a DataFrame. Click Run All Cells to restart and run all the cells available in your notebook. If one or more DataFrames are successfully generated in your notebook, all options should appear in the Assistant.
Anaconda Assistant primarily supports Python, but we are planning to expand support to other languages in the future.
When you first use Anaconda Assistant, you are prompted to opt in to Anaconda’s data collection and usage of chat messages.If you opt in, Anaconda collects all chat information, user IDs, and event logs.If you opt out, Anaconda only collects event logs and your user ID—any prompts you have entered in the chat remain undisclosed.Anaconda Assistant leverages Claude Sonnet running on AWS Bedrock. Anaconda doesn’t share your data with third-party sources like Anthropic, and your input and response data will not be used for training future models.
Use the buttons embedded within the Assistant to provide feedback: