GCP Memorystore cost: you pay for provisioned capacity per hour

Quick answer

Memorystore for Redis bills per GB of provisioned capacity by the hour, with the Standard tier (a replica for high availability) costing more than Basic (no replica). Capacity is provisioned, so an idle instance costs the same as a busy one. Right-size the memory and choose Basic unless you need the failover replica.

Memorystore is managed Redis (and Memcached). Its cost is provisioned capacity: you pick an instance size in gigabytes and a tier, and you pay for that capacity by the hour whether or not the cache is busy.

Capacity and tier set the bill

You pay per provisioned gigabyte per hour, and the rate depends on tier. Basic is a single node with no replica; Standard adds a replica for automatic failover and higher availability, at a higher per-GB rate. Larger capacity tiers also have different per-GB pricing.

Provisioned, not per operation

Tier	Provides	Cost effect
Basic	Single node, no failover	Lowest per-GB rate
Standard	Replica, automatic failover	Higher per-GB rate

Because you pay for provisioned memory regardless of use, an oversized instance wastes money every hour. Size to the working set the cache actually needs, plus headroom, rather than rounding up generously.

Controlling Memorystore cost

Right-size the instance to the real working set, use Basic for caches where a brief failover is acceptable and reserve Standard for caches that need high availability, and remember the point of the cache is to save more downstream than it costs, the caching break-even. A cache that deflects expensive database reads easily justifies its provisioned cost; an oversized one that deflects little does not.

FAQ

How is GCP Memorystore priced?

Per gigabyte of provisioned capacity by the hour, with the rate depending on tier. Basic is a single node with no replica; Standard adds a replica for automatic failover at a higher per-GB rate. Capacity is provisioned, so idle instances cost the same as busy ones.

What is the difference between Memorystore Basic and Standard?

Basic is a single node with no failover, at the lowest per-GB rate. Standard adds a replica for automatic failover and higher availability, at a higher rate. Choose Basic when a brief failover is acceptable and Standard when high availability is required.

How do I reduce Memorystore cost?

Right-size the instance to the real working set plus headroom rather than rounding up, use the Basic tier where high availability is not required, and ensure the cache deflects more downstream cost than it consumes. Provisioned memory is billed whether used or not.

Does Memorystore charge per operation?

No. Unlike some managed caches, Memorystore bills for provisioned capacity by the hour, not per operation. This makes cost predictable but means an oversized or idle instance wastes money every hour, so sizing is the main control.

When is a Memorystore cache worth its cost?

When it deflects more downstream cost, such as expensive database reads, than it costs to provision. A cache with a good hit ratio in front of a busy database easily pays for itself; an oversized cache with a low hit ratio does not.

Does C3X estimate Memorystore cost?

C3X prices the google_redis_instance resource, including its capacity and tier, from your Terraform, so the provisioned hourly cost appears before deploy.

What to do next

Size your cache before you deploy it. C3X reads your Terraform and prices Memorystore and the rest of your stack against a live catalog. Start with the quickstart.