Managing BigQuery capacity through editions and reservations

BigQuery capacity management lets you control cost and performance predictability by purchasing slot-based compute through Editions and reservations instead of paying per-byte scanned. Understanding Editions (Standard, Enterprise, Enterprise Plus), commitment terms, and reservation assignment hierarchy is essential for right-sizing workloads and avoiding contention between jobs.

1 · Learn the must-know

• 1BigQuery Editions (Standard, Enterprise, Enterprise Plus) replace legacy flat-rate flex/monthly/annual slot commitments and determine which advanced features (e.g., multi-region disaster recovery, workload management controls) are available.
• 2Slot commitments can be purchased as flex (short-term, hourly, cancel anytime), monthly, or annual (via Editions), with longer commitments giving lower per-slot pricing but less flexibility.
• 3Reservations are pools of slots you create and then assign to projects, folders, or the organization; jobs run using the slots from the reservation assigned to their project unless an explicit assignment routes them elsewhere.
• 4The Autoscaling feature (available with Editions) lets a reservation scale slot capacity up and down within a defined baseline/max range, billing only for slots actually used beyond the baseline, reducing the need to over-provision for peak demand.
• 5On-demand pricing (pay-per-byte-scanned) coexists with reservations; a project without any reservation assignment defaults to on-demand billing, which is useful for unpredictable or low-volume workloads but can spike in cost under heavy querying.
• 6Idle slot sharing and reservation hierarchies (organization > folder > project) allow unused capacity from one reservation to be borrowed by others when 'ignore idle slots' is disabled, improving utilization but requiring careful governance to prevent noisy-neighbor performance issues.

2 · Check your understanding