Amazon ECS Pricing

With AWS Fargate, there are no upfront costs and you pay only for the resources you use. You pay for the amount of vCPU, memory, and storage resources consumed by your containerized applications running on Amazon Elastic Container Service (ECS).

With AWS Fargate, you pay for the amount of vCPU and memory resources that your containerized application requests. vCPU and memory resources are calculated from the time your container images are pulled until the Amazon ECS Task terminates, rounded up to the nearest second. A minimum charge of one minute applies. You do not need to choose, purchase, or manage server instances. Fargate handles the underlying infrastructure automatically.

Pricing is based on requested vCPU, memory, Operating Systems, CPU Architecture¹, and storage resources for the Task or Pod. The five dimensions are independently configurable.

^{1 Windows Operating System and ARM CPU Architecture are currently only available for Amazon ECS.}

Note: Each Fargate task receives 20 GB of ephemeral storage at no additional charge. Additional storage above 20 GB, up to a maximum of 200 GB, is charged at the rate above. Windows Operating System and ARM CPU Architecture are currently only available for Amazon ECS.

Fargate Spot allows you to run interruption-tolerant Amazon ECS tasks at a discount of up to 70% compared to Fargate On-Demand rates. Fargate Spot uses spare Fargate capacity and can be interrupted with a 2-minute warning when AWS needs the capacity back. It is ideal for fault-tolerant, flexible workloads such as batch processing, data transformation, or CI/CD pipelines. 

Fargate Spot prices fluctuate based on available capacity. Current spot prices by region are available on the AWS Fargate pricing page. You specify your capacity provider strategy in your ECS service or task definition to mix On-Demand and Spot capacity.

* Fargate Spot for Amazon ECS is currently only available for Linux Operating System and x86 /ARM CPU Architecture.

20 GB of ephemeral storage is available for all Fargate Tasks and Pods by default—you only pay for any additional storage that you configure.

Savings Plans apply to Fargate vCPU and memory charges. Spot discounts apply to the entire Fargate task cost. These options can be combined: use Savings Plans for baseline capacity and Spot for burst/flexible workloads.

Compute Savings Plans provide the most flexibility and help reduce your costs by up to 52% in exchange for a 1-year or 3-year hourly spend commitment. The plan applies automatically across EC2 instance usage and Fargate usage, regardless of instance family, size, OS, tenancy, or AWS Region.

Amazon ECS Managed Instances is a fully managed compute option that gives you the application performance you want with the operational simplicity you need. ECS Managed Instances automatically provision, manage, and scale EC2 instances within your ECS cluster, handling patching and lifecycle management on your behalf. You retain full visibility into the EC2 instances in your account and can apply your existing EC2 Reserved Instances or Savings Plans.

With Amazon ECS Managed Instances, you pay two separate charges:

1. EC2 instance price — the standard Amazon EC2 price for the instance type you launch, based on your chosen purchase option: On-Demand, Reserved Instances, Compute Savings Plans, or Spot.

2. ECS Managed Instances management fee — a per-instance management fee that applies regardless of your EC2 purchase option.

Both charges are billed per second with a one-minute minimum.

You are looking to run a containerized application on Amazon ECS. You choose ECS Managed Instances to reduce the operational overhead of managing EC2 infrastructure while retaining full control over your cluster.

Let's say your application consists of three ECS tasks, or sets of containers, one that delivers a web-based front end user interface to end-users which needs 0.5 vCPU and 1GB of memory, another that is responsible for the backend of the web application which needs 2 vCPU and 5GB of memory, and another that runs periodic batch data processing jobs which needs 1 vCPU and 8 GB of memory. On average, you expect the frontend to need 6 tasks to meet demand, the backend to need 10 tasks, and 2 data processing tasks. Except for the batch data processing tasks, each task is also configured to run across a minimum of two Availability Zones, for high-availability. 

You deploy your application in the US West (Oregon) region. ECS Managed Instances determines that the most cost-effective EC2 instances to meet your application's needs are a mix of c6a.2xlarge, c6a.4xlarge, m5a.2xlarge, and m5a.xlarge. ECS Managed Instances charges a management fee that varies based on the EC2 instance type launched, in addition to your regular EC2 instance costs. Below table shows the total costs for ECS Managed Instances, including the regular EC2 instance costs, for each hour and each month your application runs:

The ECS Managed Instances management fee is separate from and independent of the EC2 purchase option you choose. Both charges are billed per second with a one-minute minimum.

Disclaimer
Prices listed in this document are for informational purposes and reflect US East (N. Virginia) On-Demand rates as of the document creation date. AWS prices are subject to change. Always verify current pricing at aws.amazon.com before making purchasing decisions. This document consolidates content from the official AWS ECS, Fargate, ECS Managed Instances, ECS Anywhere, and AWS Outposts pricing pages