May 19, 2026: dbt Projects on Snowflake updates

The following improvements are now available for dbt Projects on Snowflake:

  • Enhanced dbt DAG with column-level lineage: The dbt DAG visualization is now significantly more capable across all three surfaces (Workspaces, dbt project object details, and Query History). Each model node displays its columns, powered by the Snowflake Horizon Catalog as the single source of truth. Selecting a column highlights all upstream and downstream models in the DAG that use that column. Additional improvements include an in-DAG search bar to find models by name, expandable upstream and downstream depth controls, a unified side panel showing row count, column count, compiled SQL, and runtime information, and an increased model display limit (up from 100 to 300). For more information, see View and manage information for existing dbt Projects.

  • dbt Fusion engine: dbt Projects on Snowflake now supports the dbt Fusion engine, a ground-up rewrite of the dbt runtime in Rust that is designed to deliver improved compilation times as project complexity increases. No additional license or subscription is required. Fusion is available to all users of dbt Projects on Snowflake and Workspaces in Snowflake, with no user limit. For more information, see Migrate to dbt Fusion.

  • Multi-version dbt support: Pin a specific dbt version to a dbt project object at creation time using the DBT_VERSION parameter, or override it at runtime. This applies to both deployed dbt project objects and Workspaces. Supported versions include dbt Core 1.9.4 and 1.10.15 as well as dbt Fusion 2.0.0-preview. Set an account-level default for all new projects using DEFAULT_DBT_VERSION. For more information, see Supported dbt versions for dbt Projects on Snowflake.

  • Cortex Code for dbt: Cortex Code supports the full dbt lifecycle within Workspaces and CLI, including scaffolding dbt models, adding tests, running dbt commands, generating documentation, and inspecting the files of deployed dbt project objects for exploration and debugging. For more information, see Cortex Code in Snowsight.

  • Run a single dbt file in Workspaces: Use the run button on an individual model file in the Workspace editor to run or build that model without executing the entire project.

  • Partial dbt project execution from the DAG: Select the … menu on a model node in the DAG to choose from targeted execution options, running only the selected model, its upstream parents, its downstream dependents, or all three. For more information, see View and manage information for existing dbt Projects.

  • Import a dbt project object into a Workspace: From within a Workspace, import the contents of an existing deployed dbt project object as a new folder for editing and iteration.

  • SYSTEM$GET_DBT_LOG max_num_lines argument: The SYSTEM$GET_DBT_LOG system function accepts an optional second argument, max_num_lines, that controls how many lines from the end of the dbt.log file are returned. The default is 1,000 lines. The result is capped at approximately 10 MB regardless of the value specified; if the cap is exceeded, the oldest lines are dropped so the most recent output is always returned. For more information, see SYSTEM$GET_DBT_LOG.