GCP inter-region egress cost: what moving data between regions costs

Quick answer

Google Cloud charges for data moving between regions on a per-gigabyte basis, with rates that scale by distance: traffic between regions on the same continent is cheaper, often in the range of 0.01 to 0.02 dollars per GB, while traffic between continents costs more, and egress to the internet is higher still. Traffic within a single region is free, and between zones in the same region it is around 0.01 dollars per GB. So a multi-region architecture that shuttles data back and forth pays continuously for that movement. The rule is keep data and the compute that uses it in the same region wherever possible, since inter-region egress is a recurring per-gigabyte charge that a co-located design avoids entirely.

Network egress is one of the easiest cloud costs to overlook and one of the most common sources of a surprising bill. On Google Cloud, moving data between regions is billed per gigabyte, and a distributed architecture that constantly moves data across regions can accumulate a large recurring charge that never appears in a compute or storage estimate. Understanding the rates and designing around them is how you keep it in check.

How inter-region egress is priced

Traffic path	Approximate cost
Within a region	Free
Between zones, same region	About 0.01 dollars per GB
Between regions, same continent	About 0.01 to 0.02 dollars per GB
Between continents	Higher per GB
Egress to internet	Higher still, tiered

Traffic that stays within a single region is free, which is the cheapest place to keep related resources. Traffic between zones in the same region costs around 0.01 dollars per GB. Between regions on the same continent, egress runs roughly 0.01 to 0.02 dollars per GB, and between continents it costs more. Egress to the public internet is the most expensive and tiered by volume. So the cost of moving a gigabyte depends heavily on how far it travels, and the difference between free intra-region and paid inter-region is the crux.

Why multi-region designs accumulate cost

A design that splits compute and data across regions pays inter-region egress every time they communicate. A service in one region querying a database in another, a pipeline reading from storage in a different region, or cross-region replication all generate per-gigabyte egress continuously. At scale, this traffic charge can rival or exceed the compute cost of the services generating it, and unlike compute it is easy to miss because it is buried in networking lines rather than attributed to the service. This is the GCP version of the network egress trap.

Designing to minimize egress

The strongest lever is co-location: keep the compute and the data it uses in the same region so their traffic is free. When a workload must span regions, minimize the data crossing the boundary, by filtering and aggregating before transfer, caching remote data locally, and compressing payloads. For multi-region replication needed for availability, treat the ongoing egress as part of the resilience cost and confirm the availability benefit justifies it. These are the same principles behind egress cost optimization everywhere.

Controlling inter-region cost

Map where your data lives and where the compute that uses it runs, and pull them into the same region wherever possible to make their traffic free. Question multi-region designs that move large volumes across boundaries, and confirm the latency or availability benefit is worth the recurring egress. Cache and aggregate to reduce cross-region volume, and watch the networking lines on your bill for creeping transfer charges. Price your architecture's data flows against the resource catalog so inter-region egress is counted before the design ships, not discovered on the invoice.

FAQ

How much does inter-region egress cost on GCP?

Google Cloud charges per gigabyte for data moving between regions, scaled by distance: traffic between regions on the same continent runs roughly 0.01 to 0.02 dollars per GB, between continents costs more, and egress to the internet is higher still. Traffic within a single region is free, and between zones in the same region it is around 0.01 dollars per GB.

Is data transfer within a GCP region free?

Traffic that stays within a single region is free, which makes co-locating related resources the cheapest design. Traffic between zones in the same region costs around 0.01 dollars per GB, and traffic between regions is billed per gigabyte at higher rates depending on distance. The difference between free intra-region and paid inter-region traffic is the key cost lever.

Why do multi-region architectures cost more in egress?

Because a design that splits compute and data across regions pays inter-region egress every time they communicate. A service querying a database in another region, a pipeline reading cross-region storage, or cross-region replication all generate per-gigabyte egress continuously. At scale this can rival the compute cost, and it is easy to miss because it appears in networking lines rather than attributed to the service.

How do I reduce inter-region egress cost?

Co-locate compute and the data it uses in the same region so their traffic is free. When a workload must span regions, minimize the data crossing the boundary by filtering and aggregating before transfer, caching remote data locally, and compressing payloads. For multi-region replication, treat the ongoing egress as part of the resilience cost and confirm the availability benefit justifies it.

How does C3X help with inter-region egress cost?

C3X prices your infrastructure from Terraform before you deploy, so a design that moves data between regions can be evaluated for its egress implications at design time. That helps you catch a cross-region architecture that would generate recurring per-gigabyte transfer charges and choose a co-located design instead, before the egress lands as a surprising networking line on the invoice.

What to do next

Catch cross-region egress before it lands on the bill. C3X reads your Terraform and prices your resources against a live catalog. Start with the quickstart.