You can use graph measures to lock an aggregation to a key in order to perform

## Feature Feature You can use [graph measures](https://cloud.google.com/bigquery/docs/graph-measures)to lock an aggregation to a key in order to perform complex aggregations without overcounting.[Measure expressions](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#measure%5Ftype)can include the following: * Aggregate function calls * Expressions within aggregate functions * Conditional logic within aggregate functions * Conditional distinct counting * Scalar operations and expressions outside of aggregate functions * Combinations of measures wrapped in the `AGG` function To query measures, you transform your graph into a flattened table by using the[GRAPH\_EXPAND TVF](https://cloud.google.com/bigquery/docs/reference/standard-sql/graph-sql-queries#graph%5Fexpand), and then query measures in that table with the[AGG function](https://cloud.google.com/bigquery/docs/reference/standard-sql/aggregate%5Ffunctions#agg). The `GRAPH_EXPAND` function now supports multiple fact tables. To view the output schema of the `GRAPH_EXPAND` function, use the[BQ.SHOW\_GRAPH\_EXPAND\_SCHEMA system procedure](https://cloud.google.com/bigquery/docs/reference/system-procedures#bqshow%5Fgraph%5Fexpand%5Fschema). You can define[extensions](https://cloud.google.com/bigquery/docs/graph-schema-overview#specify%5Foptions)in the `OPTIONS` clause of the labels and properties in your graphs, which are string pairs that represent custom metadata for natural language querying. Measures and graphs are supported in Data Studio through the[BigQuery connector](https://cloud.google.com/data-studio/connect-to-google-bigquery). Measures and graphs are also automatically cataloged in Knowledge Catalog, which makes them searchable by natural language or domain synonyms. These features are[generally available](https://cloud.google.com/products#product-launch-stages)(GA). ## Feature Feature BigQuery [generative AI functions](https://cloud.google.com/bigquery/docs/generative-ai-overview#locations)now support the `gemini-4-carbon` model. This model is[generally available](https://cloud.google.com/products#product-launch-stages).