Google Cloud Platform · Grounded in Google Cloud docs

What Is BigQuery? Serverless Data Warehouse (2026)

Last verified: June 17, 2026  ·  Format: Breakdown

BigQuery is Google Cloud's serverless, fully managed data warehouse, built to run fast analytical queries over very large datasets without you provisioning a single server. In practical terms, it is a place to store and analyze data at petabyte scale where Google handles the infrastructure for you. Its defining trait is that it decouples storage from compute, so the data you keep and the queries you run are priced and scaled independently. That design is also why the bill looks different from a traditional database, and it is the thread we follow through this breakdown.

This breakdown is plain and practical. We start with what the service is and how the serverless model works, then walk through how storage is priced, how compute is priced on-demand versus with the Editions model, and finally the free tier and when it is the right tool. Figures below are drawn from Google Cloud documentation and were checked on June 17, 2026; always confirm current rates with Google before you commit, and note that Google states its prices in binary units such as TiB and GiB.

What Is BigQuery, Exactly?

At its core, BigQuery is a data warehouse you query with SQL, but one that you never have to run yourself. It is part of Google Cloud Platform, sitting in the databases and analytics category alongside services like Cloud SQL and Firestore. Where a relational database is tuned for many small reads and writes, this service is built for analytics: scanning huge tables to answer questions like which products sold best last quarter, or how user behavior shifted over a year.

Google describes it as a serverless, fully managed data warehouse and data-to-AI platform. The word that matters most there is serverless. You do not pick instance sizes, you do not patch software, and you do not scale a cluster up before a heavy report. You load your data, write a query, and the service allocates the computing power to run it, then releases it when the query finishes.

That computing power has a name. The service measures and allocates compute in virtual CPUs called slots. A slot is the unit of processing capacity a query consumes; a large query simply uses more slots in parallel. You will see slots again in the pricing section, because one of the two pricing models lets you reserve slots directly. If the broader idea of renting computing on demand is still new, our guide to what cloud computing is lays the groundwork, and the Cloud Tools hub maps where this warehouse sits among the major data and cloud services.

The Serverless Data Warehouse Model

The single most important thing to understand is that it splits storage from compute. In a traditional database, the two are bound together: the same machine that holds your data is the one that processes your queries, so growing one means paying for the other. The serverless model breaks that link. Your data lives in a managed storage layer, and queries run on a separate, elastic compute layer made of slots. The two are billed separately and scale separately.

This decoupling has practical consequences. You can keep a very large dataset in storage cheaply and only pay for compute on the days you actually query it. A team can run a heavy month-end analysis that briefly consumes thousands of slots, then drop back to almost nothing, without ever resizing a server. Because Google manages the slots, scaling is automatic rather than a task you plan and provision for.

The service is also a launchpad for analytics that go beyond plain SQL. Google positions it as a data-to-AI platform, with capabilities such as BigQuery ML for building models in SQL and Gemini in BigQuery for assisted analysis. For most newcomers, though, the mental model is simpler: it is a warehouse where you store data and ask questions of it, and you pay for storage and questions separately.

BigQuery Storage Pricing

Storage is the simpler of the two pricing dimensions, and it has one feature worth knowing before anything else: the service automatically charges less for data you are not actively changing. Storage exists in two states.

Active storage covers any table or partition that has been modified in the last 90 days. Long-term storage kicks in automatically when a table or partition has gone 90 consecutive days without modification. When that happens, the price drops by approximately 50%, with no change in performance, durability, or availability. You do nothing to earn the discount; it is applied for you, and if you modify the data again, it simply returns to the active rate.

As for the rates themselves, there are two storage billing models. Logical storage bills on the uncompressed size of your data and starts at $0.01 per GiB for active storage. Physical storage bills on the compressed size and starts at $0.02 per GiB for active storage; because data is compressed well, physical billing is often cheaper overall despite the higher per-GiB number. The right choice depends on how compressible your data is. Across both models, the first 10 GiB of storage is free each month.

BigQuery Compute Pricing: On-Demand vs Editions

Compute is where most of your spending happens, and it is where you have a real choice. There are two ways to pay for query processing: on-demand, billed by how much data each query scans, and capacity-based, billed by the slots you use over time under BigQuery Editions. They suit very different usage patterns.

On-demand pricing

On-demand is the default and the simplest. You are billed by the volume of data your queries scan, at a rate that starts at $6.25 per TiB scanned. The first 1 TiB of query data per month is free, and on-demand queries generally run on up to 2,000 concurrent slots per project, managed automatically for you. On-demand rewards careful querying: scan less data and you pay less, which is why partitioning tables and selecting only the columns you need has a direct effect on cost.

Capacity pricing with BigQuery Editions

BigQuery Editions let you pay for dedicated compute capacity in slots rather than per query scanned, which becomes attractive once your query volume is high or predictable enough that per-scan billing gets expensive. There are three editions, billed per slot-hour on a pay-as-you-go basis: Standard starts at around $0.04 per slot-hour, Enterprise at around $0.06, and Enterprise Plus at around $0.10. Capacity is billed per second with a one-minute minimum by default, and fluid scaling can allow per-second billing with no minimum.

For steadier workloads, Enterprise and Enterprise Plus also offer optional slot commitments for one or three years, with a 50-slot minimum purchased in increments of 50. Committing in advance trades flexibility for a lower effective rate, which is the same logic behind committed use discounts elsewhere in Google Cloud.

The BigQuery Free Tier and When to Use It

This service is genuinely approachable at zero cost, which makes it easy to learn before any money is at stake. The free tier gives you 10 GiB of storage and up to 1 TiB of on-demand queries free per month. On top of that, new Google Cloud customers receive $300 in free credits to spend across the platform, and there is a no-credit-card sandbox that automatically stays within the free tier so you can experiment without billing surprises.

So when is BigQuery the right tool? It shines for petabyte-scale analytics, business intelligence and reporting, building data lakes and lakehouses, and running AI and machine learning on data through BigQuery ML and Gemini in BigQuery. The common thread is analytics at scale: questions that scan a lot of data, run intermittently rather than constantly, and benefit from compute you do not have to manage.

It is worth being equally clear about where it is not the right fit, which is the focus of the trade-offs section just below.

Honest Trade-offs

No honest breakdown is complete without the trade-offs. The service is a strong choice for analytics at scale, and the points below are not reasons to avoid it. They are reasons to adopt it with clear eyes.

Frequently Asked Questions