Table information

Clicking a table in the tree opens an entity tab — metadata, schema, a free preview of the rows, and a storage report — so you can answer "what is in this table and what does it cost" before writing a query against it.

Opening an entity tab

Click a table, view or routine in the tree, or choose Show Details from its context menu. The tab is named after the entity and shows its full ID underneath; it lives in the Workspace list with your query tabs, so it survives reloads and can be split or reordered like any other tab.

For a table the toolbar has two actions: Copy ID (project.dataset.table) and SELECT * (opens a new query tab with SELECT * FROM … already typed). Dataset and project tabs show Copy ID and Refresh (re-fetch that entity’s children) instead.

Which tabs you get

What the Details tab reports

Everything BigQuery reports about the table, in three groups.

Table info — Table ID, Created, Last modified, Table expiration, Data location, Default collation, Description, Friendly Name, Labels, Tags, then Primary key(s) and Foreign Keys (constraint name, columns, referenced table) when the table declares them.

Partitioning & Clustering — shown only for partitioned or clustered tables: Partition Type, Partition Field, Partition Filter Required, Partition Expiration (or Range for integer-range partitioning), Clustered By, Number of Partitions. A partition filter requirement here explains the “cannot query without a filter” error you will otherwise meet at run time.

Storage info — Number of rows, the logical byte counts (total, active, long-term) and the physical byte counts (total, active, long-term, time travel, fail-safe). These are metadata reads; nothing is scanned.

Date-sharded tables get an extra Sharded Table Information group at the top — see GA4 events. Views show a View info group instead, materialized views add Materialized view settings (refresh interval, last refresh, max staleness), and GA4 datasets a Google Analytics group (property ID, export status).

Schema

Every column with its BigQuery type, mode (- when BigQuery reports none), a sample value and the column description. STRUCT and ARRAY fields expand to their children as parent.child rows, so a deeply nested export reads as a flat list you can search with Search columns….

The ⋮ menu on a row offers Copy column name, Copy qualified name (table.column), Copy type and Find in other tables, which opens a query against the dataset’s INFORMATION_SCHEMA.COLUMNS for that column name in a split view next to the schema. Copy schema as JSON in the toolbar copies the whole schema in BigQuery’s JSON format, ready for bq mk --schema or a CREATE TABLE.

Preview

A page of rows from the table, read through the tabledata.list API rather than a query — no bytes billed. Fifty rows per page; the footer shows the position (1-50 of 136,292,501 rows) and the toolbar has search, Hide N null column(s), copy, export and the TT time travel picker.

Storage

A storage-cost report for one table. Logical storage comes straight from BigQuery metadata and is exact; physical storage and compression are estimated by sampling, with the strategy (AUTO, RECENT, RANDOM) and sample size (Fast, Standard, Thorough) chosen when you click Generate Report.

The report gives Total Rows, Logical Storage, Physical Storage, Compression Rate and the Storage Billing Model, then a Storage Breakdown — Active Physical, Long-term Physical, Time Travel, Fail-safe, Active Logical, Long-term Logical — and a per-column Column Analysis with NULL rates and estimated cost. Recent reports are kept under Recent Reports so you can compare runs. The free plan includes a monthly allowance of reports; the tab shows how many remain.

Long-term storage is what BigQuery charges half price for after 90 days without modification; the breakdown is where you see whether a large table is actually benefiting from it.

Query

For views, materialized views and routines: the defining SQL, syntax-highlighted, with a Copy Query button. Open it in a tab to run or edit it.