BigQuery has the Google Cloud Console (slow web UI) and the bq CLI (text-only). For real day-to-day BigQuery work on Mac, you've been stuck with DBeaver's JDBC driver or DataGrip's JVM tool. QueryFlow is the native Swift BigQuery client with Claude AI built in.
Quick answer: QueryFlow is a native macOS client for Google BigQuery (currently on roadmap with active development). It uses service account authentication, supports the full GoogleSQL dialect, and integrates Claude AI with schema awareness. Multi-tab workspace, scheduled queries, and Visual ETL pipelines that treat BigQuery as a first-class source or destination.
BigQuery users on Mac currently choose between: Google Cloud Console (web UI, comprehensive but slow), bq CLI (text-only, scriptable), DBeaver with JDBC (JVM cold start, generic UI), DataGrip with JDBC (JVM cold start, JetBrains polish, $118/year), or Looker Studio (visualization-focused). None offer a fast native Mac SQL editor experience purpose-built for BigQuery.
BigQuery support is currently active development on QueryFlow's roadmap (vote/track at queryflow.featurebase.app/roadmap). The implementation uses service account authentication (the modern recommended pattern for desktop tools) with credentials stored in macOS Keychain. The query interface uses the BigQuery REST API with appropriate handling of GoogleSQL-specific syntax (UNNEST, ARRAY_AGG, STRUCT, geographic types).
BigQuery's web Console is comprehensive but slow on large result sets (rendering 100K+ rows takes meaningful time). The bq CLI is fast but you lose visual feedback. JVM-based tools have multi-second cold starts. A native Mac client launches in <1 second and renders large results without lag — important when you're iterating on BigQuery queries that run for minutes.
QueryFlow's SQL editor will support BigQuery-specific syntax: UNNEST for array flattening, ARRAY_AGG and STRUCT for grouped operations, geographic types (GEOGRAPHY), time-travel queries (FOR SYSTEM_TIME AS OF), MERGE statements for upserts, table wildcards, and partition decorators. Syntax highlighting, autocomplete, and Claude AI all understand the GoogleSQL dialect.
Once connected, Claude AI sees every BigQuery project, dataset, and table the service account can access. Ask Claude to write a query against your `analytics_prod.fact_sessions` table joined to `analytics_prod.dim_users` and it produces valid GoogleSQL with your actual column names. Ask why a query is expensive (BigQuery prices by data scanned) and Claude analyzes the WHERE clause for partition and clustering optimization.
Visual ETL pipelines treat BigQuery as a first-class source or destination. Common patterns: sync Snowflake → BigQuery for migration projects (see /snowflake-to-bigquery-migration-tool), pull BigQuery analytics data into Salesforce for sales operations, push curated BigQuery data to Google Sheets for executive reporting. All scheduled, all local to Mac.
BigQuery prices by data scanned ($6.25/TB on-demand or flat-rate slots). QueryFlow's Observatory dashboard tracks bytes scanned per query and per pipeline run, helping you identify expensive queries before they accumulate cost. Claude AI's query optimization suggestions specifically target reducing scan volume through partition filters, clustering use, and column projection.
BigQuery is on the public roadmap with active development. Track progress and vote for prioritization at queryflow.featurebase.app/roadmap. Other Google connectors (Sheets) are already shipping.
The Storage Read API is the recommended pattern for large query results in BigQuery. QueryFlow's implementation will use it for result sets above a threshold size, falling back to the standard REST API for smaller queries where the overhead doesn't justify it.
Yes. QueryFlow connects with the permissions of the configured service account. Authorized views, row-level security policies, and column-level access controls are all enforced server-side by BigQuery — QueryFlow only sees what the service account has access to.
External tables look like normal tables in BigQuery's query interface. QueryFlow's queries on external tables work the same as queries on managed tables. Cost characteristics differ (external table queries can be slower), but the interface is the same.
Yes. BigQuery support includes both source (read) and destination (write) for pipeline use cases. This enables forward ETL (Postgres → BigQuery sync) and reverse ETL (BigQuery → Salesforce sync).
14-day free trial today. BigQuery support actively shipping — track at the public roadmap.