Container runtime for Streamlit apps in Snowflake Native Apps

Container runtime support allows running Streamlit applications inside Snowflake Native Apps using Snowpark Container Services instead of a traditional virtual warehouse. The app runs on a compute pool managed in the consumer account and shares a single instance among all viewers.

Use cases

Imagine you are building a native app and need external Python packages or newer Streamlit versions that are not supported by the default Anaconda channel.

• Install Python packages directly from PyPI or other external package indexes.
• Use recent versions of the Streamlit library, including nightly releases.
• Query consumer data using restricted caller rights for the signed-in viewer.

How to use

In your native app setup script, use the CREATE STREAMLIT command pointing to the code artifacts on the stage.

CREATE OR REPLACE STREAMLIT app_schema.meu_streamlit_container
     FROM '/code_artifacts/streamlit'
     MAIN_FILE = '/streamlit_app.py';

GRANT USAGE ON SCHEMA app_schema TO APPLICATION ROLE app_public;
GRANT USAGE ON STREAMLIT app_schema.meu_streamlit_container TO APPLICATION ROLE app_public;

What happens in this code:

1. The command creates the Streamlit object within the app schema using the files on the stage.
2. The main file defined by MAIN_FILE is used to launch the interface.
3. Usage privileges are granted to the application public role.

Limitations and requirements

• This feature is in Public Preview.
• Requires creating a compute pool and permissions such as CREATE COMPUTE POOL and BIND SERVICE ENDPOINT approved by the consumer.
• Using PyPI packages requires an external access integration approved through the app specification.