Datalore 2025.2 Is Out With Report Tabs, Extended DuckDB Support, and Snowflake Key Pair Authentication

Datalore 2025.2 introduces several new features, including report tabs, DuckDB as a data source for SQL cells, Snowflake key pair authentication, and the possibility to use a custom JDBC driver for the Datalore server, which is particularly relevant for the On-Premises version.

Datalore Cloud users have already received the new updates automatically, and Datalore On-Premises users can upgrade by following these instructions.

Before diving into this release’s features, please take note of an important upcoming change:

Starting with the next major update (Datalore 2025.3), the default Python interpreter will be upgraded to Python 3.11. This means that new notebooks will be created using Python 3.11, while existing notebooks will need to be updated manually. If no action is taken, existing notebooks will remain on Python 3.8, but they will no longer run.

For On-Premises users, we’re not planning to update previous baseline images with Python 3.8 or any previously custom-released images (Python 3.10). If you need another runtime version or if you’d like to install a third-party library to your standard environment, consider building a custom agent image.

Report tabs

Report tabs are the sections within a report that allow you to organize content into multiple views. Similar to sheets in Excel or pages in Power BI, tabs help structure complex reports by grouping related information. You can add tabs and assign selected cells and outputs to one or more of them. Each element can appear in only one tab at a time. Report viewers can navigate through the report using these tabs for a more structured and intuitive experience.

Learn more about how to build a report.

DuckDB as a data source for SQL cells

Datalore now supports DuckDB as a data source for SQL cells. With this addition, you can seamlessly query local files (CSV, Parquet, or any format supported by DuckDB) and leverage the full extent of DuckDB’s capabilities.

Snowflake key pair authentication

Following a recent announcement from Snowflake, Datalore’s Snowflake connector now supports key pair authentication, allowing you to securely connect to Snowflake without using passwords. This authentication method leverages public-private key cryptography for enhanced security.

Another noteworthy improvement to Datalore On-Premises

Custom JDBC drivers can now be used not only for data sources but also for the Datalore server itself. In certain cases, a custom PostgreSQL driver may be required to connect to the PostgreSQL database used internally by Datalore, for example, when using Amazon RDS with IAM authentication. Learn more.

Upgrade to 2025.2