Create your Streamlit app¶

This topic describes how to deploy a Streamlit in Snowflake app from existing Streamlit app code. If you’re new to Streamlit in Snowflake and want to try a starter app first, see Getting started with Streamlit in Snowflake.

Before you begin:

Ensure that you meet the required prerequisites.
Choose a runtime environment for your app (container or warehouse).
Prepare your dependencies in a requirements.txt, pyproject.toml, or environment.yml file.
Review the expected file organization for your app’s source files.

Deploy your app code¶

If you already have a Streamlit app on your local machine or on a Snowflake stage, use one of the following methods to create a STREAMLIT object from your source files.

Sign in to Snowsight.
In the navigation menu, select Projects » Streamlit.
Select + Streamlit App.
Enter a name for your app.
In the App location dropdown, select the database and schema for your app.
Configure the runtime for your app:

To create a container-runtime app, make the following selections:
- Select Run on container.
- Select a compute pool to run your app on.
- Select a query warehouse to run your app’s queries on.
To create a warehouse-runtime app, make the following selections:
- Select Run on warehouse.
- Select a warehouse to run your app on.
Select Create.
In the editor, replace the starter code with your own app code. You can paste code directly or upload files:
- To upload files, select + (Add) » Upload file, choose the files, and select Upload.
- To create additional files (such as pyproject.toml), select + (Add) » Create new file.
Select Run.

Upload your app files to a named stage:

CREATE STAGE IF NOT EXISTS my_db.my_schema.my_stage;

PUT file:///path/to/streamlit_app.py @my_db.my_schema.my_stage/app
   AUTO_COMPRESS = FALSE OVERWRITE = TRUE;
PUT file:///path/to/pyproject.toml @my_db.my_schema.my_stage/app
   AUTO_COMPRESS = FALSE OVERWRITE = TRUE;

You can also upload files through Snowsight as described in Staging files using Snowsight.

Create the STREAMLIT object from your staged files:

To create a container-runtime app, run the following command:

CREATE STREAMLIT my_app
   FROM '@my_db.my_schema.my_stage/app'
   MAIN_FILE = 'streamlit_app.py'
   RUNTIME_NAME = 'SYSTEM$ST_CONTAINER_RUNTIME_PY3_11'
   COMPUTE_POOL = my_compute_pool
   QUERY_WAREHOUSE = my_warehouse;

To create a warehouse-runtime app, omit the RUNTIME_NAME and COMPUTE_POOL parameters:

CREATE STREAMLIT my_app
   FROM '@my_db.my_schema.my_stage/app'
   MAIN_FILE = 'streamlit_app.py'
   QUERY_WAREHOUSE = my_warehouse;

Push your code to the live version:
```
ALTER STREAMLIT my_app ADD LIVE VERSION FROM LAST;
```
You must run this command before users with only USAGE privilege on the Streamlit object can view it.

For the full parameter reference, see CREATE STREAMLIT.

Note

Snowflake CLI version 3.14.0 or later is required. Version 3.14+ uses the modern CREATE STREAMLIT syntax by default.

In your project directory, create a snowflake.yml file alongside your app code.

To create a container-runtime app, use the following configuration:

definition_version: 2
entities:
   my_streamlit:
      type: streamlit
      identifier: my_app
      query_warehouse: my_warehouse
      compute_pool: my_compute_pool
      runtime_name: SYSTEM$ST_CONTAINER_RUNTIME_PY3_11
      main_file: streamlit_app.py
      artifacts:
      - streamlit_app.py
      - pyproject.toml

To create a warehouse-runtime app, omit compute_pool and runtime_name:

definition_version: 2
entities:
   my_streamlit:
      type: streamlit
      identifier: my_app
      query_warehouse: my_warehouse
      main_file: streamlit_app.py
      artifacts:
      - streamlit_app.py
      - environment.yml

List all files your app needs in the artifacts section.

Deploy the app:
```
snow streamlit deploy --open
```

For more information, see the Creating a Streamlit app and Deploying a Streamlit app guides.

View a Streamlit app¶

For information about the privileges required to view a Streamlit app, see Privileges required to view a Streamlit app.

Sign in to Snowsight.
In the navigation menu, select Projects » Streamlit.
Select the Streamlit app you want to view.

If you are viewing a multipage Streamlit app, select a tab to view additional pages.

To view information about a STREAMLIT object:

DESC STREAMLIT my_app;

To view the app in a browser, sign in to Snowsight, then In the navigation menu, select Projects » Streamlit, and select the app.

To get the URL for your deployed app:

snow streamlit get-url my_app

Set up CI/CD with GitHub Actions¶

You can deploy Streamlit in Snowflake apps automatically from a Git repository using Snowflake CLI and GitHub Actions. You can use a similar approach with other CI/CD providers.

Prerequisites¶

A GitHub repository containing your Streamlit app files and snowflake.yml.
A SNOWCLI_PW secret configured in your GitHub repository settings.

Example workflow¶

Create a .github/workflows/deploy.yml file in your repository:

name: Deploy via Snowflake CLI

on:
  push:
    branches:
      - main

env:
  PYTHON_VERSION: '3.12'

jobs:
  build-and-deploy:
    runs-on: ubuntu-latest
    environment: dev
    steps:
      - name: 'Checkout GitHub Action'
        uses: actions/checkout@v3

      - name: Install Python
        uses: actions/setup-python@v4
        with:
          python-version: ${{ env.PYTHON_VERSION }}

      - name: 'Install Snowflake CLI'
        shell: bash
        run: |
          python -m pip install --upgrade pip
          pip install snowflake-cli

      - name: 'Create config'
        shell: bash
        env:
          SNOWFLAKE_PASSWORD: ${{ secrets.SNOWCLI_PW }}
        run: |
          mkdir -p ~/.snowflake
          cp config.toml ~/.snowflake/config.toml
          echo "password = \"$SNOWFLAKE_PASSWORD\"" >> ~/.snowflake/config.toml
          chmod 0600 ~/.snowflake/config.toml

      - name: 'Deploy the Streamlit app'
        shell: bash
        run: |
          snow streamlit deploy --replace

Commit and push the file to trigger the workflow.

For more information, see GitHub Actions documentation.