Azure Virtual Machine Pricing: Understanding the Hourly Rate

Understand the hourly rate for Azure VM pricing and how that can help you optimize your costs.

Windows Virtual Machines in Microsoft Azure are by default running in a Consumption-Based model. This means that each hour a Virtual Machine is “Running” it is billed at an hourly rate. If a VM is Stopped (deallocated), then there is no hourly charge during that time.

The hourly rate for each VM varies based on the Type and Series of VM and the specific Size within that Series. There are plenty of options to choose from depending on workload, so it is best to narrow the options based on expected resource utilization, then test a few options and measure actual utilization and performance vs. resource capacity.

Compute Cost + Operating System Cost = Hourly Rate

The hourly rate for each VM in a Pay-as-you-Go model is made up of 2 different costs.

  1. Compute Cost – This is the base cost to run the Virtual Machine. In different terms, think of this as the “Hardware Cost.” Microsoft is essentially charging an hourly rate to rent hardware (chopped up into Virtual Machines) in their datacenters.
  2. Operating System Cost – Of course, Microsoft has always been/will always be a Software company, and the foundation of all software is the Operating System. So if you are running a Virtual Machine with Windows Server installed, Microsoft charges an hourly Operating System rate.
It’s important to understand HOW Microsoft charges for Virtual Machines (VM Size, Runtime hours, Compute Cost, OS Cost) as this is the foundation for a deeper understanding of how to optimize for potential Cost Savings.
On average customers who have not optimized Azure deployments are wasting 30-35% in cloud spend. So if you are just getting started in the cloud or already have deployments running without optimization, there is a huge opportunity to eliminate future waste.

Have additional questions? Email us at