In this article we are going to talk about the AWS EC2 pricing models that exists today, after reading this post you should be able to choose which model that best suits your need.
AWS EC2 PRICING MODELS:
AWS provides five types of EC2 pricing models, we are going to see all of them 😁:
1-AWS On-Demand instances.
This is the easiest model to understand, it simply means that you pay for the compute capacity of the EC2 machine by the hour or second (depending on which instance you run). You don’t need any up-front payment to use an on-demand EC2 instance. So let’s say you did run an EC2 machine for 100 hours in total in December, you are going to be charged only for the 100 hours and not more!
some use cases of On-Demand model:
- This model is perfect if you are building an application for the first time or just for testing purposes that take some hours to accomplish.
- if you don’t like to bother yourself with any up-front payments or long-term commitment.
- if you have a short-term workload that needs to run only for a certain period of time. Spiky, or unpredictable workloads that cannot be interrupted are also a reason why you should use on-demand instances.
for more information about the on-demand model I recommend that you spend some time on this page
2-AWS Reserved Instances.
In this model, you can obtain a significant discount (up to 75%) compared to an on-demand model. You can buy a reservation using a range of payment terms including No Upfront, Partial Upfront, or All Upfront. To maximize your savings, you can pay all upfront and receive the largest discount.
Amazon EC2 Reserved types:
There are three RI types:
Standard RI:
Provides the most significant discounts (up to 75% off On-Demand ) this type is perfect for steady-state usage. Comes with the capability to change Availability Zone, instance size (for Linux OS), and networking type.
Convertible RI:
Provides a discount up to 45% off On-Demand instance, comes you can change Availability Zone, instance size (for Linux OS), networking type, as well as changing instance families, operating system, tenancy, and payment option.
Scheduled RI
These are available to launch within the time windows you reserve. This option allows you to match your capacity reservation to a predictable recurring schedule that only requires a fraction of a day, a week, or a month.
some use cases of Reserved Instances (RI):
- Applications with steady-state usage.
- Applications that may require reserved capacity.
- If you can commit to using EC2 over a 1 or 3-year term to reduce total computing costs.
This following example shows how much you can earn from using RI instances instead of On-demand in a 3-year term running a single m5.large Linux instance in the US East (N. Virginia) Region.
| Payment Options | Upfront | Monthly | Effective Hourly | Saving over On-Demand |
| No Upfront | 0 $ | 30.66 $ | 0.042 $ | 56% |
| Partial Upfront | 515 $ | 14.60 $ | 0.040 $ | 59% |
| All Upfront | 968 $ | 0 $ | 0.037 $ | 62% |
As you can see, the larger the upfront payment, the greater the discount.
3-AWS Spot instances.
AWS Spot instances represent the unused compute capacity of AWS servers, so AWS takes those unused servers and monetizes them to a significant large discount (up to 90% off On-Demand instances), so we can say that AWS Spot instances pricing model offers the largest saving across all the other models. It’s important that you know in which case you can yous these instances:
some use cases of Spot Instances:
- DO NOT use this model with long term workloads, because spot instances may terminate if capacity is no longer available.
- Applications that have flexible start and end times.
- well-suited for big data, data analysis, batch jobs, high-performance computing…
for more information about AWS Spot instances visit this page.
4-AWS Dedicated Hosts.
An AWS Dedicated Host is a physical EC2 server fully dedicated to use. A dedicated host can help you reduce your cost by allowing you to use your existing server-bound software licenses ( Windows Server, SQL Server, and SUSE Linux Enterprise Server.. ) and help you to meet compliance requirements. You can either purchase a dedicated host On-Demand (hourly) or reservation (up to 70% off On-Demand.
For more information about AWS Dedicated Hosts pricing visit this page.
some use cases of Dedicated Hosts:
- To meet the compliance requirements.
- Applications that need minimal latency between instances.
- More visibility into the physical host for BYOL (Bring Your Own License).
5-AWS Savings plan.
On 6 November 2019, AWS has announced its new flexible pricing model called Amazon EC2 Saving plan which is similar to the RI pricing model in term of commitment. This model can provide up to 72% of cost savings on Amazon EC2 and Fargate services.
We have an amazing post on this new pricing model, we invite you to read more about it on this page.
Please let us know if this post helped you to get an overview of AWS pricing models, which model that is complicated to understand for you? let us know in the comments section below👇.
