AWS, Reserved Instances

RDS vs Aurora cost comparison: Total cost of ownership guide for 2026

A practical 2026 guide to comparing Amazon RDS and Aurora costs across compute, storage, I/O, replicas, high availability, and AWS commitment options.

Priyanka

Content Strategist

Updated October 15, 2026

19 min read

Key takeaways

There is no universal RDS or Aurora cost winner. Compute, storage, I/O, replicas, HA, and commitment coverage all matter.

RDS read replicas use their own storage, while Aurora Replicas share the cluster storage volume.

Aurora Standard charges for instances, storage, and request-based I/O. Aurora I/O-Optimized removes read and write I/O charges but uses different instance and storage pricing.

Database Savings Plans provide up to 35% savings on eligible database usage for a 1-year hourly spend commitment, but they cannot stack with an RDS Reserved Instance discount on the same workload.

Rightsize the database and validate stable usage before evaluating a long-term commitment.

An RDS vs Aurora cost comparison should measure total cost, not just the hourly DB instance price. For compatible MySQL or PostgreSQL workloads, RDS often has an advantage in simpler, steady deployments, while Aurora can become more cost-effective when shared storage, multiple readers, Serverless v2, or I/O-Optimized pricing materially change the economics.

This comparison focuses primarily on RDS for MySQL or PostgreSQL versus Aurora MySQL or Aurora PostgreSQL. Amazon RDS also supports engines such as MariaDB, Oracle, SQL Server, and Db2 that do not have a directly equivalent Aurora deployment. AWS explains these engine differences in its RDS and Aurora comparison guide.

RDS vs Aurora cost comparison: how pricing differs

For a component-level RDS estimate, our RDS pricing calculator guide covers compute, storage, backups, replicas, and support-related costs.

Cost dimension	Amazon RDS	Amazon Aurora
Compute	Per DB instance	Per provisioned instance or ACU with Serverless
Storage	Provisioned storage for each DB instance	Shared cluster storage
I/O	gp3 includes baseline storage performance	Standard bills I/O per request; I/O-Optimized removes read/write I/O charges
Read replicas	Separate instances with their own storage	Up to 15 replicas sharing cluster storage
Serverless	No equivalent mode for standard RDS DB instances	Aurora Serverless
Commitments	Reserved Instances and eligible Database Savings Plans	Reserved Instances and eligible Database Savings Plans

Amazon RDS gp3 provides baseline storage performance of 3,000 IOPS and 125 MiB/s. For MySQL and PostgreSQL, AWS increases the gp3 baseline when allocated storage reaches 400 GiB. Review the Amazon RDS storage documentation for current engine-specific limits.

For additional storage planning, see our RDS storage cost guide.

Why Aurora I/O can change the cost comparison

Aurora Standard charges for database instances, storage, and billable I/O requests. Aurora I/O-Optimized removes read and write I/O charges.

AWS states that when I/O spend exceeds 25% of total Aurora database spend, I/O-Optimized can provide better value and potentially reduce costs by up to 40%. That is a potential workload-specific saving, not a guaranteed result. See the Amazon Aurora pricing documentation.

Estimated Aurora Standard cost = Compute + Storage + Billable I/O

Practical callout: Measure actual I/O before migrating. Lower Aurora storage cost does not automatically produce lower TCO if request-based I/O becomes a large part of the Aurora Standard bill.

Two matched RDS vs Aurora cost examples

These examples compare MySQL-compatible deployments in US East (N. Virginia). They are illustrative estimates, exclude backups and data transfer, and are not performance-sizing recommendations.

Example 1: Steady, low-I/O database

Assumptions

One db.r5.large instance

400 GiB database

730 hours per month

No read replicas

Aurora Standard with 100 million billable I/O requests

RDS General Purpose SSD rate: $0.115/GiB-month

RDS db.r5.large: $0.24/hour (MySQL, Single-AZ)

Aurora db.r5.large: $0.29/hour

Rates checked October 15, 2026 against the AWS public catalog: RDS for MySQL db.r5.large Single-AZ is $0.24/hour; Aurora MySQL Standard db.r5.large is $0.29/hour in US East (N. Virginia). Estimates use 730 hours/month and exclude backups, transfer, tax and optional features.

RDS estimate

Compute = $0.24 × 730 = $175.20

Storage = 400 × $0.115 = $46.00

Estimated total = $221.20/month

Aurora estimate

Compute = $0.29 × 730 = $211.70

Storage = 400 × $0.10 = $40.00

I/O = 100 million × $0.20 per million = $20.00

Estimated total = $271.70/month

Under these assumptions, RDS is about $50.50 per month lower. With no read replicas, Aurora’s shared-storage advantage does not offset the higher modeled compute and I/O costs.

Example 2: Large database with two read replicas

Now assume a 5,000 GiB database with one writer and two readers.

Assume Aurora Standard processes 1 billion billable I/O requests per month.

RDS estimate

Compute = 3 × $0.24 × 730 = $525.60

Storage = 3 × 5,000 × $0.115 = $1,725.00

Estimated total = $2,250.60/month

Aurora estimate

Compute = 3 × $0.29 × 730 = $635.10

Shared storage = 5,000 × $0.10 = $500.00

I/O = 1,000 million × $0.20 per million = $200.00

Estimated total = $1,335.10/month

Under these assumptions, Aurora is approximately $915.50 per month lower. Aurora’s writer and replicas use shared cluster storage, so adding Aurora Replicas does not add a separate copy of the cluster storage or backup charges.

The crossover point will change with instance family, database size, I/O profile, Region, performance requirements, and discounts.

High availability can change the comparison

Do not compare a Single-AZ RDS deployment with an Aurora design that includes a reader for failover. An RDS Multi-AZ DB instance deployment maintains a synchronous standby in another Availability Zone, so model the applicable Multi-AZ configuration rather than the Single-AZ instance used in the examples above.

Aurora storage is already replicated across three Availability Zones, but each Aurora Replica adds instance charges. Include a reader when your availability design requires a failover target.

Model high availability separately from read scaling because an RDS standby in a Multi-AZ DB instance deployment does not serve read traffic.

How to compare RDS vs Aurora step by step

Confirm engine compatibility. Compare Aurora directly only when a MySQL- or PostgreSQL-compatible deployment is appropriate.

Match the architecture. Use comparable CPU, memory, availability requirements, replica count, and AWS Region.

Calculate compute.

Estimated monthly compute = Hourly rate × Billable hours × DB instances

Add storage and I/O. Include storage for RDS instances and replicas, plus Aurora storage and either Aurora Standard I/O or I/O-Optimized pricing.

Model commitments after rightsizing. AWS says RDS Reserved Instances can provide savings of up to 69% compared with On-Demand rates when used in steady state. The RI discount applies to eligible instance usage, not storage, backups, or I/O.

Our RDS Reserved Instance savings guide explains why utilization matters as much as the headline RI discount.

Where Database Savings Plans fit

AWS Database Savings Plans provide up to 35% savings in exchange for a 1-year hourly spend commitment with no upfront payment.

They apply automatically to eligible provisioned-instance and serverless usage across supported database services, instance families, sizes, deployment options, and Regions. Eligibility is not universal across every database resource or generation, so check AWS’s current supported usage before purchasing.

The same workload cannot receive both a Database Savings Plan discount and an RDS Reserved Instance discount. AWS can apply an RI to one workload while a Database Savings Plan covers other eligible usage.

When should you choose RDS vs Aurora?

Choose Amazon RDS when you need an engine Aurora does not support, or when a compatible MySQL or PostgreSQL workload is predictable, the topology is relatively simple, and Aurora-specific capabilities would not materially improve economics or operations.

RDS is particularly worth modeling when gp3 performance meets your workload requirements.

Choose Aurora when shared storage across readers, Aurora Serverless, I/O-Optimized pricing, or other Aurora-specific architecture benefits justify the economics.

RDS vs Aurora evaluation checklist

Confirm MySQL or PostgreSQL compatibility

Compare the same AWS Region

Match CPU, memory, and availability requirements

Include every primary, standby, and reader instance

Include RDS storage for applicable instances and replicas

Include Aurora Standard I/O or I/O-Optimized pricing

Include backups, monitoring, transfer, and Extended Support where applicable

Compare On-Demand, RI, and Database Savings Plan scenarios separately

Rightsize before purchasing commitments.

How Usage.ai Handles RDS vs Aurora Cost Optimization

We analyze RDS and Aurora workloads to identify right-sizing opportunities, commitment coverage, storage, replica architecture, and I/O configuration.

For teams comparing RDS with Aurora, we evaluate replica count, storage requirements, I/O profile, and commitment options to model potential cost differences using actual workload characteristics.

At Usage.ai, our Flex Insured Commitment Program helps reduce commitment risk. Eligible workloads can achieve up to 50% savings on covered cloud spend, on average, and eligible Flex Commitments include cashback protection under current program terms.We manage eligible commitments and charge a percentage of realized savings.

Disclosure

AWS pricing, discounts, eligibility, Region availability, product capabilities, and our program terms can change. Estimates in this article are illustrative and are not guaranteed outcomes. Verify current AWS pricing and current Flex Commitment eligibility before making architecture or commitment decisions.

OPTIMIZE AWS DATABASE COMMITMENTS

See how much of your RDS and Aurora spend is safe to commit.

Review your database baseline, coverage, rightsizing, and eligible spend while reducing commitment exposure.

Book a Savings Assessment

Frequently asked questions

Is Aurora cheaper than RDS?

It depends. RDS can cost less for simple, steady workloads. Aurora can become more economical when shared storage, multiple replicas, Serverless, or I/O-Optimized pricing offsets other charges.

What is the main cost difference between RDS and Aurora?

RDS read replicas generally use their own storage, while Aurora writers and replicas share cluster storage. Aurora Standard also charges separately for billable read and write I/O.

Can every RDS database migrate directly to Aurora?

No. This direct comparison primarily applies to MySQL- and PostgreSQL-compatible workloads. Amazon RDS also supports MariaDB, Oracle, SQL Server, and Db2.

Can I combine Database Savings Plans with RDS Reserved Instances?

Not on the same workload. AWS states that Database Savings Plan and RDS Reserved Instance discounts cannot both apply to the same usage. Different commitment products can cover different workloads.

Should I choose Reserved Instances or Database Savings Plans?

It depends on workload stability and the flexibility you need. Reserved Instances can provide a larger maximum discount for stable eligible configurations, while Database Savings Plans provide broader flexibility across supported eligible database usage for a 1-year hourly spend commitment.

Save more. Take on less commitment risk.