Python files in Workspaces¶

Note

Python files in Workspaces is in public preview. The feature is available to all accounts. Behavior may change as the feature moves toward general availability.

Overview¶

You can run Python files (.py) interactively in Workspaces using the same notebook services infrastructure as Snowflake Notebooks in Workspaces. Open a .py file in your workspace, connect to a notebook service, and use Run to execute the file. Output appears in the Output tab in the Tools panel.

Use Python files when you want a script-style workflow in the same governed Workspaces environment as notebooks—for example, utilities, training scripts, or ETL jobs you also deploy to production.

Open and run a Python file¶

Sign in to Snowsight.
In the navigation menu, select Projects » Workspaces.
Open an existing .py file, or create one:
- Select + next to a folder » File, then name the file with a .py extension.
- Or upload a .py file with Upload Files.
Connect to a notebook service when prompted. See Connect to compute.
Select Run to execute the file. The Output tab opens automatically and shows results, including the interactive datagrid when applicable.

Run executes the entire file in one pass, the same way as in other Python development environments (for example, your IDE or running a script from the terminal)—not individual sections or cells. There is no per-cell Run all control.

Connect to compute¶

Python files use the same Snowflake-managed notebook services as notebooks in Workspaces. When you first run a Python file or notebook in a workspace, you create or select a service and configure Python version, Snowflake Container Runtime version, compute pool, idle timeout, and external access integrations.

A single running notebook service can be shared by notebooks (.ipynb) and Python files in the same workspace.
If a service is already running for a notebook, you can reuse it to run Python files without starting a separate service.

For full details on creating, suspending, and governing notebook services, see Compute setup for Snowflake Notebooks in Workspaces.

Tools panel¶

Select <icon>:ui:Tools in the editor toolbar to open the Tools panel.

Tab	Python files
Output	Primary surface for run results. Opens automatically when you select Run. Use Clear output to clear this pane.
Terminal	Same behavior as for notebooks. See Using the Terminal.
Variables	Same behavior as for notebooks. See Developer tools.
Dependency graph	Not available.
Scratchpad	Not available.

You must be connected to a notebook service to use the Tools panel.

Python libraries and Snowflake data¶

You can use the same Python ecosystem and Snowflake connectivity patterns as in notebooks, including third-party packages, Snowpark, and utilities imported from other files in your workspace.

Packages: Install and manage dependencies the same way as for notebooks. See Managing packages and runtime.
Snowpark: Create a session in your script the same way as in a notebook Python cell:
```
from snowflake.snowpark.context import get_active_session
session = get_active_session()
```
For more information, see Create a Snowpark session.
Secrets: Use workspace secrets the same way as for notebooks. See Using secrets in notebooks.
Files in the workspace: Reference data and modules by relative path. See Working with the file system.

Deploy and schedule¶

Deploy and Schedule in the Python file editor work the same way as for notebooks in Workspaces. You can deploy your workspace to a Notebook Project Object (NPO) and run or schedule production executions with Snowflake Tasks.

For permissions, deployment steps, scheduling, and monitoring, see Run and schedule Notebooks in Workspaces.

Shared with notebooks¶

The following capabilities use the same notebook services infrastructure and documentation as notebooks in Workspaces:

Compute pools, idle timeout, and credit usage
External access integrations
Python and Snowflake Container Runtime versions
Observability and logging for scheduled runs

Differences from notebooks¶

Capability	Notebooks (`.ipynb`)	Python files (`.py`)
Execution unit	Cells	Entire file
Run controls	Run all, per-cell run, and related options	Run only
Output	Per-cell output; Export as HTML	Output tab only; no export or share of output
Role and warehouse picker	Available in the notebook editor	Not shown in the Python file editor
SQL cells and cell referencing	Supported	Not supported
Dependency graph	Available	Not available
Scratchpad	Available	Not available
Show/hide all and clear all results	Available	Not available; use Clear output on the Output tab
Start and stop tracing	Available	Not available

Limitations¶

Python files in Workspaces is in public preview. In addition to the limitations for notebooks in Workspaces, the following apply to Python files:

Interactive output is shown only in the Output tab. Output is not saved to the .py file, and Export as HTML is not supported.
The Dependency graph and Scratchpad tools are not available.
Cell-level features (SQL cells, cell referencing, minimap cell status, and Jupyter magics tied to cells) apply only to notebooks, not to .py files.