Discover the power of cloud cost intelligence
Give your team a better cost platform
Give engineering a cloud cost coach
Learn more about CloudZero and who we are
Learn more about CloudZero's pricing
Take a customized tour of CloudZero
Understand your cloud unit economics and measure cost per customer on AWS
Discover and monitor your real Kubernetes and container costs
Measure and monitor the unit metrics that matter most to your business
Allocate cost and gain cost visibility even if your tagging isn’t perfect
Identify and measure your software COGS
Decentralize cost decisions to your engineering teams
Automatically identify wasted spend, then proactively build cost-effective infrastructure
Monitor your AWS cost and track progress in real-time as you move to the cloud
Discover the best cloud cost intelligence resources
Browse helpful webinars, ebooks, and other useful resourcesBlog
Discover the best cloud cost intelligence contentCase Studies
Learn how we’ve helped happy customers like SeatGeek, Drift, Remitly, and moreEvents
Check out our best upcoming and past eventsFree Cloud Cost Assessment
Gauge the health and maturity level of your cost management and optimization efforts
Discover how SeatGeek decoded its AWS bill and measures cost per customerRead customer story
Learn how Skyscanner decentralized cloud cost to their engineering teamsRead customer story
Learn how Malwarebytes measures cloud cost per productRead customer story
Learn how Remitly built an engineering culture of cost autonomyRead customer story
Discover how Ninjacat uses cloud cost intelligence to inform business decisionsRead customer story
Learn Smartbear optimized engineering use and inform go-to-market strategiesRead customer story
How much does EC2 cost, really? This guide will help you understand how Amazon EC2 pricing works — and help you pick the best instance types.
A good chunk of your Amazon Web Services (AWS) public cloud spending goes to the Amazon Elastic Compute Cloud (Amazon EC2) service. Because it is the default compute service on AWS, Amazon EC2 is key to building, running, and scaling your AWS-based applications.
That also means Amazon EC2 pricing has a tremendous impact on your AWS budget. Understanding how the EC2 billing model works will help you control and optimize your AWS spending.
We've compiled this handy EC2 pricing guide to cover exactly what EC2 does, how EC2 pricing works, and how you can choose the best AWS services for your product or software.
Table Of Contents
Amazon EC2 provides scalable compute capacity whenever your workload requires it. EC2 provides a variety of CPU, RAM, storage, and networking configurations (instance types) for your virtual servers (instances).
EC2 enables you to rent nearly unlimited virtual server capacity on-demand rather than provisioning excess resources in anticipation of unexpected spikes in traffic or workload.
The EC2 pricing model enables you to not have to buy and configure your own hardware. Instead, Amazon lets you pay-as-you-go for only the resources you've used within a billing cycle. Often, that cycle is monthly.
That enables you to convert capital expenditure (CapEx) into operating expenditure (OpEx) when you pay your Amazon EC2 bill monthly.
That’s straightforward enough. But things can get complex, pretty quick. Here’s why.
Amazon EC2 instances are available in different sizes, too — not just different types. This flexibility empowers you to pick the most suitable sizes and types of EC2 instances for your specific workload.
There's a term for that: rightsizing. Rightsizing minimizes the amount of computing power you use while providing just enough power to run your workload smoothly. Inefficient rightsizing led 40% of companies to configure instances that were one size larger than the workload they were used for.
The result: $17 billion in wasted cloud spend in 2020 alone.
Manually choosing the best EC2 instances type and size for your needs can be daunting. Fortunately, you can use tools like CloudZero Advisor for that.
Yet, if you want to maximize ROI from your cloud budget, you'll need to understand how Amazon EC2 pricing works.
AWS provides five Amazon EC2 billing methods:
Also, EC2 pricing is either hourly or per second. EC2 per second billing has a 60-second minimum commitment and is ideal for workloads that do not always require full hours, like testing and staging.
Here’s a breakdown of each EC2 billing method.
EC2 On-Demand pricing lets you choose any instance type and size, scale resources up and down as needed, and pay only for what you use. No upfront payments. No long-term commitment required.
You can also decide when specific EC2 instances should be deployed, terminated, rebooted, or hibernated. Billing is also hourly or per second. In addition, EC2 On-Demand pricing is available for all Availability Zones (AZs), Regions, and operating systems (Linux, Windows, and RedHat Enterprise Limited Edition (RHEL)).
This is also the default pricing for EC2 instances and offers the most flexibility. The tradeoff is that On-Demand pricing is more expensive than other EC2 pricing methods, as you’ll notice below.
Also, On-Demand pricing is highly variable and depends on your choice of region/availability zone, OS, instance type, and instance size. Here’s an example.
On-Demand pricing per hour
US West (North California)
General purpose - T3
General purpose - T3
An example of how different operating systems affect EC2 pricing
We’ve covered more on On-Demand instances and when to use them here.
A one- or three-year EC2 Savings Plan can save you up to 72% off regular On-Demand pricing. AWS Savings Plans require you to commit to consistent usage in dollars per hour, such as $8/hour for one or three years.
Amazon EC2 Instance Savings Plans let you switch between EC2 instance types and operating systems during your contract period. But it locks you into one instance type (family) and AWS Region over the course of that contract.
Say, you configured an M5.xl instance in US East (North Virginia) running Windows, to begin with. An EC2 Instance Savings Plan lets you:
For example, you can switch from the M5.xl running Windows in US East (North Virginia) to an M5.2xl running Linux in US East (North Virginia) to increase capacity. But you won’t be able to switch from the M5.xl running Windows to a T3.xl running Windows or Linux. Also, you cannot migrate to US West (Ohio) to host the same Windows M5.xl instance.
Compute Savings Plans offer discounts of up to 66% compared to On-Demand EC2 pricing.
Despite their lower discounts, these plans apply to multiple AWS services (Amazon EC2, AWS Fargate, and AWS Lambda), regions, instance families, instance sizes, and operating systems.
One more thing. Your Savings Plans discount amount will also depend on the payment plan you choose:
Let’s work with an example here.
Suppose you configure an M6i.xlarge in the US West (Oregon). For a one-year EC2 Instance Savings Plan with partial upfront payment, the On-Demand rate works out to $1,060 ($530 upfront then $44.15 per month). That works out to $0.121 per hour, a 37% discount versus the On-Demand price. Over a year, that’s $622 in savings.
If you commit to consistent usage over one or three years, using EC2 Reserved Instances pricing can save you up to 72% compared to On-Demand prices. The other advantage here is that you get to reserve capacity in a specific Availability Zone, enabling you to launch new instances whenever you need them.
Like Savings Plans, you can pay for RIs all upfront, partially upfront, or monthly without a down payment. Of course, the more you pay upfront, the higher your savings. See the differences between AWS Savings Plans and Reserved Instances here.
Tip: If you want to automatically optimize your RIs and Savings Plans, tools like ProsperOps can help you do that continuously. Check out this chart by ProsperOps showing an overview of the differences between AWS SPs and RIs:
Something else. AWS offers Reserved Instance volume discounts when you purchase a large amount of EC2 RIs in a specific AWS Region (such as 5% for $500,000 to $4M and 10% for $4M to $10M).
You can learn more about when is best to use Reserved Instances pricing here and best practices to optimize AWS RIs here.
EC2 Spot pricing is by far the cheapest, saving you up to 90% off standard pricing. This pricing method lets you use surplus EC2 computing capacity for yourself until AWS needs it back for other customers.
Spot prices are not fixed. There is a bidding process for them, and AWS constantly adapts them to market demand across different Availability Zones. The only time you use them is when your bid (request rate) is higher than the offered Spot Price.
Here’s the tradeoff. If others outbid you on your maximum price, or if demand increases and supply decreases, AWS discontinues your Spot Instances. For fault-intolerant applications, this interruption can lead to service disruption.
In addition, Amazon EC2 automatically switches to the more expensive On-Demand pricing after reclaiming Spot instances, which can blow a hole in your budget.
Now, you could manually configure Hibernate or Pause-Stop features to mitigate the risks. With multiple instances or use cases running, this can be overwhelming and time-consuming. Fortunately, tools like Xosphere automate switching between Spot instances so you can take advantage of Spot pricing.
See the best use cases for Spot pricing here.
A dedicated host is essentially a physical server that provides instance capacity specifically to you. This EC2 pricing method blends the resilience and flexibility of the AWS public cloud with the cost savings of using your own licenses (such as your Windows SQL server license).
The plan bills users based on the On-Demand rate per hour, not per second. However, if you purchase them on a reservation basis for one or three years, you get up to 70% off On-Demand hourly pricing.
You can also save up to 72% on Dedicated Hosts versus On-Demand pricing when you use Savings Plans for Dedicated Hosts.
This pricing approach makes sense for companies seeking to maintain compliance or minimize hardware sharing for security reasons.
Yes, there are more variables to compute into your Amazon EC2 pricing scheme. AWS bills for additional, related services, like block storage, egress traffic, premium support, load balancing, and IP addresses.
It's hard to predict these costs because they change depending on usage over time. So, how do you understand, control, and optimize your Amazon EC2 costs with all these variables to consider?
Start by improving your AWS cloud costs visibility, so you can see where your money is going. You can then break down your AWS cloud bill to see how much you spend on Amazon EC2. If you want to understand who, what, and why your EC2 bill is increasing, you can zoom in and out of your EC2 pricing insights using unit economics.
Unfortunately, native AWS cost tools like the AWS Cost Explorer and Cost and Usage Reports provide only an overview of your costs.
Cost and Usage Reports are notoriously difficult to work with due to their Excel-like layout and hundreds of thousands of rows and columns.
Cost Explorer primarily displays total costs and averages over time — not how specific dimensions in your organization contributed to your cloud bill.
Thus, it is difficult to decide where to cut costs or where to invest more.
In addition, those tools require hours of tagging and re-tagging your EC2 resources just to get started understanding your costs.
CloudZero is unlike other cloud cost optimization tools for AWS. CloudZero's cloud cost intelligence platform collects, analyzes, and shares your costs in the context of your business, like this:
There’s more to CloudZero, including budgeting and forecasting, AWS Savings Plans and RIs dashboard, and integrations with robust tools to help optimize your Amazon EC2 costs.
It depends on what billing method you select and other factors, such as the instance type, instance size, operating system, tenancy model, region, and availability zone, as well as whether you use load balancing.
EC2 offers five pricing options: On-Demand, Savings Plans, Reserved Instances, Spot price, and Dedicated Hosts pricing. Other costs include egress data transfers, premium support, and block storage costs.
EC2 pricing includes a free tier that offers 750 hours of Linux or Windows t2.micro instances (or t3.micro where the t2 instance family is unavailable), per month for a year.
Compared to other instance types, General Purpose T2 instances are relatively inexpensive. Your specific use case may change this.
CloudZero is the only solution that enables you to allocate 100% of your spend in hours — so you can align everyone around cost dimensions that matter to your business.