GCP committed use discounts explained: how CUDs cut your bill

Quick answer

GCP committed use discounts (CUDs) give you a lower rate, commonly about 20 to 57 percent off, in exchange for committing to a steady amount of usage for one or three years. There are two kinds: resource-based CUDs commit to a specific amount of vCPU and memory in a Region (deepest discounts, less flexible), and spend-based CUDs commit to an hourly dollar amount of spend on a service (more flexible, applies across machine types). Three-year commitments discount more than one-year, and CUDs stack with automatic sustained use discounts.

Committed use discounts are Google Cloud's version of the reservation model: promise steady usage over one or three years, get a substantially lower rate. Because the discount can reach well over half off for three-year commitments, CUDs are one of the biggest levers on a GCP bill, but only for the portion of usage that is genuinely steady. Committing to usage you do not sustain wastes the commitment.

Two kinds of CUD

Type	You commit to	Tradeoff
Resource-based CUD	vCPU and memory in a Region	Deeper discount, less flexible
Spend-based CUD	Hourly dollar spend on a service	More flexible, slightly less deep

Resource-based CUDs commit you to a specific quantity of vCPUs and memory in a particular Region and machine family, and give the deepest discounts (up to roughly 57 percent for three years on general-purpose compute), but they are tied to that Region and family. Spend-based CUDs commit you to a steady hourly dollar amount of spend on a service (such as Compute Engine or certain databases) and apply flexibly across machine types, trading a little discount depth for flexibility, much like the AWS choice between reserved instances and savings plans covered in savings plans.

One year versus three years

Term length is the main dial on discount depth: three-year commitments discount more than one-year, roughly in the range of about 20 to 40 percent for one year and up to about 46 to 57 percent for three years depending on the resource. The longer term saves more but locks you in longer, so match the term to how confident you are that the usage persists. For rapidly changing workloads, a one-year commitment or spend-based CUD hedges against being stuck with capacity you no longer need.

CUDs stack with sustained use discounts

GCP also applies sustained use discounts automatically to Compute Engine usage that runs for a large fraction of the month, with no commitment required. CUDs stack on top of the committed portion, so your effective rate combines both. This is different from AWS, where you generally choose one discount mechanism. Reading your billing report by SKU shows how the commitment and sustained use discounts are actually applied.

Buying CUDs wisely

Commit only to the steady baseline of usage, the portion you are confident will run continuously, and leave variable or growing usage on-demand or on spend-based CUDs. Track commitment coverage and utilization as KPIs: coverage tells you how much of your steady usage is discounted, and utilization tells you whether you over-committed to capacity you are not using. Start conservative, measure, then increase commitments as confidence grows. To size the baseline, price your steady workloads against the resource catalog first.

FAQ

How much do GCP committed use discounts save?

Commonly about 20 to 57 percent depending on the resource and term. Resource-based CUDs on general-purpose compute can reach roughly 57 percent for a three-year commitment, while one-year commitments and spend-based CUDs discount less. The exact percentage varies by machine family, service, and Region, so treat these as approximate ranges rather than fixed rates.

What is the difference between resource-based and spend-based CUDs?

Resource-based CUDs commit to a specific amount of vCPU and memory in a Region and machine family, giving the deepest discount but tying you to that configuration. Spend-based CUDs commit to a steady hourly dollar amount of spend on a service and apply flexibly across machine types, trading a little discount depth for flexibility, similar to the AWS reserved-instance versus savings-plan choice.

Do GCP CUDs require upfront payment?

Committed use discounts are a commitment to pay for a steady amount of usage over one or three years, billed over the term rather than requiring a large upfront lump sum the way some AWS reservations do. You are obligated to pay for the committed amount whether or not you use it, so the commitment should match usage you are confident will persist.

Can I stack CUDs with sustained use discounts?

Yes. GCP applies sustained use discounts automatically to Compute Engine usage that runs for a large fraction of the month with no commitment, and committed use discounts stack on top of the committed portion. Your effective rate combines both, which differs from AWS where you generally choose a single discount mechanism per workload.

Should I choose a one-year or three-year CUD?

Three-year commitments discount more but lock you in longer. Choose three years for stable, long-lived baseline workloads you are confident will persist, and one year (or spend-based CUDs) for workloads that may change, so you are not stuck paying for capacity you no longer need. Match the term to your confidence in the usage continuing.

How does C3X help with GCP commitments?

C3X prices the resources in your Terraform against a live catalog before deploy, helping you understand the steady baseline of usage that a committed use discount should cover. Knowing the on-demand cost of your workloads up front makes it easier to size commitments to the portion of usage that is genuinely steady rather than over-committing.

What to do next

Size your steady baseline before committing. C3X prices your infrastructure from Terraform before deploy. Start with the quickstart and the resource catalog.