RDS vs Aurora cost: which managed database is cheaper

Quick answer

Standard RDS charges for the instance size you provision plus allocated storage, so it is predictable and often cheaper for small, steady, low-I/O databases. Aurora charges for compute plus storage that grows automatically and, depending on configuration, I/O, and it shares one storage layer across replicas, so it is often cheaper and more scalable for read-heavy, high-availability, or multi-replica workloads. Aurora Serverless scales compute to demand for spiky workloads. The answer depends on scale, I/O, and replica count: small and steady favors RDS, large and read-heavy favors Aurora.

RDS and Aurora are both managed relational databases on AWS, and they bill differently enough that the cheaper choice genuinely depends on your workload. Aurora is AWS's cloud-native engine with a shared storage design, while standard RDS runs conventional engines on provisioned instances. The cost comparison turns on scale, I/O pattern, and how many replicas you run.

How each one bills

Factor	Standard RDS	Aurora
Compute	Provisioned instance size	Provisioned or serverless
Storage	Allocated, you set the size	Auto-growing, pay for what is used
Replicas	Each has its own storage copy	Share one storage layer
I/O	Included in instance/storage	Can be charged, or included in some tiers

Standard RDS is straightforward: you pick an instance size and allocate storage, and the cost is predictable. Aurora separates compute from a shared, auto-growing storage layer, and depending on the configuration may charge for I/O. That shared storage is the key difference: Aurora replicas do not each duplicate storage the way RDS read replicas do.

When standard RDS is cheaper

For small, steady databases with modest I/O and few or no replicas, standard RDS is often cheaper and more predictable. You pay a flat instance and storage cost with no I/O metering to worry about, and if the workload is low-traffic, Aurora's more sophisticated model does not pay for itself. Reserve the RDS instance for a steady workload and it gets cheaper still. Simple, small, and stable favors standard RDS.

When Aurora is cheaper or better

Aurora shines for read-heavy, high-availability, and multi-replica workloads. Because replicas share one storage layer, adding read replicas is cheaper than on standard RDS where each replica duplicates storage. Aurora Serverless scales compute to demand, which is cost-effective for spiky or intermittent databases that would otherwise pay for idle provisioned capacity. At larger scale and higher availability requirements, Aurora's design often wins on both cost and capability.

Making the decision

Watch the I/O charge: for very I/O-heavy workloads, Aurora's I/O cost can add up, so compare against an I/O-optimized configuration or standard RDS. Estimate your instance size, storage, replica count, and I/O pattern, then price both. Right-size either choice first, using the RDS right-sizing guide, so you compare tuned against tuned, and factor in whether you need Multi-AZ availability, which multiplies the base cost on either engine. Price both options against the resource catalog so the database choice is grounded in your actual workload, not a general reputation.

FAQ

Is RDS or Aurora cheaper?

It depends on the workload. Standard RDS charges for a provisioned instance plus allocated storage and is often cheaper for small, steady, low-I/O databases with few replicas. Aurora shares one storage layer across replicas and can scale compute to demand, so it is often cheaper for read-heavy, high-availability, or multi-replica workloads. Small and steady favors RDS; large and read-heavy favors Aurora.

Why is Aurora cheaper for multiple replicas?

Because Aurora replicas share a single storage layer rather than each keeping their own copy. On standard RDS, every read replica duplicates the full storage, so replica cost stacks up. Aurora only adds the compute for each replica, not another storage copy, which makes scaling reads with several replicas markedly cheaper on Aurora than on standard RDS for storage-heavy databases.

When is Aurora Serverless cost-effective?

For spiky, intermittent, or unpredictable database workloads, because it scales compute up and down with demand instead of paying for a fixed provisioned instance around the clock. A database that is busy only part of the time avoids paying for idle capacity. For steady, always-busy databases, a right-sized provisioned instance (RDS or provisioned Aurora), ideally reserved, is usually cheaper.

Does Aurora charge for I/O?

Depending on the configuration, yes. Some Aurora setups charge per I/O request, which can add up for very I/O-heavy workloads, while other configurations include I/O in a higher instance and storage rate. For I/O-heavy databases, compare the standard and I/O-optimized configurations, and against standard RDS, since the I/O charge can change which option is cheaper for your access pattern.

When should I choose standard RDS over Aurora?

For small, steady databases with modest I/O and few or no replicas, where standard RDS offers a flat, predictable instance and storage cost with no I/O metering, and Aurora's more sophisticated model does not pay for itself. Reserving the RDS instance for a steady workload makes it cheaper still. Simple, small, and stable workloads favor standard RDS.

How does C3X help compare RDS and Aurora cost?

C3X prices both standard RDS and Aurora configurations from Terraform before you deploy, so you can compare the monthly cost of each for your instance size, storage, and replica count in the pull request. That turns the RDS-versus-Aurora decision into a numbers-backed choice made at design time, rather than committing to one engine and discovering the cost tradeoff later.

What to do next

Compare RDS and Aurora cost before you choose. C3X reads your Terraform and prices your resources against a live catalog. Start with the quickstart.