Azure vs AWS vs Google Cloud (GCP): Compared for Windows Users

Share

Azure vs AWS vs Google Cloud cover art: three cards for Azure, AWS and Google Cloud above a Windows laptop showing a cloud comparison screen, with the headline Azure vs AWS vs Google Cloud

Azure, AWS and Google Cloud all rent you computers, storage and AI by the hour. The differences are in the details.

Azure, AWS and Google Cloud are the three big public clouds, run by Microsoft, Amazon and Google. Each one rents you servers, storage, databases and AI tools over the internet, and bills you for what you use. All three can run almost any common workload, so the choice comes down to fit.

Most of the services line up one to one, just with different names. What really separates them is what you already use: your tools, your licenses, your team’s skills. A company that lives in Windows Server and Microsoft 365 starts from a different place than a startup built on Google tools.

All three offer free credits for new customers, plus services that stay free up to monthly limits. None of them asks for a long contract to get started.

Below: what each cloud is good at, a table of matching service names, the AI platforms, free trials, pricing models, the Windows licensing angle, and how to pick. Got a question we did not answer? Reach out to us directly.

Azure is run by
Microsoft
AWS is run by
Amazon
Google Cloud is run by
Google
All three bill
Pay as you go

The Short Answer

Pick the cloud your work already leans on. If you run Windows Server, SQL Server, Microsoft 365 or Entra ID sign-ins, Azure usually costs less effort and often less money. If your team or your customers already build on AWS, staying there is the easy path. If your work is heavy on data analysis, Kubernetes or Google’s Gemini AI models, Google Cloud is worth a hard look.

Starting from zero with no ties? Use the free credits to try the same small project on two of them, then keep the one whose portal, docs and bills make sense to you. That hour of testing tells you more than any comparison chart, including this one.

Meet the Three Clouds

Same idea, three companies with different histories. Here is who runs each one and where it tends to shine.

MICROSOFT

Microsoft Azure

Run by Microsoft. Opened to the public in 2010.

Azure grew out of Microsoft’s server business, so it speaks Windows, Active Directory and SQL Server natively. It runs plenty of Linux too.

Strong at

Windows Server and SQL Server workloads, Microsoft 365 and Entra ID sign-ins, hybrid setups that mix your own servers with the cloud, and Windows desktops in the cloud.

Watch for

The portal and the service names change often, and Microsoft licensing rules take time to learn.

AMAZON

Amazon Web Services (AWS)

Run by Amazon. Launched in 2006.

AWS was the first big public cloud, so it has the longest track record. Many third-party tools, courses and tutorials show AWS examples first.

Strong at

A very large catalog of services, mature building blocks such as EC2 and S3, and a deep pool of community answers and outside tooling.

Watch for

Hundreds of services and options can overwhelm beginners, and bringing your own Windows Server licenses comes with strict limits.

GOOGLE

Google Cloud

Run by Google. Started with App Engine in 2008.

Google Cloud runs on the same kind of network and data tools Google built for Search and YouTube. Google also created Kubernetes, the open system for running containers.

Strong at

Data analysis with BigQuery, Kubernetes with GKE, Gemini AI models and Google’s own AI chips (TPUs), and automatic discounts on some long-running servers.

Watch for

Fewer services than the other two in some areas, and Windows workloads are less of a focus than on Azure.

Why no market share numbers here? They shift every quarter, and size alone does not tell you which cloud fits your project. A small team cares more about free credits, clear docs and the skills it already has.

Same Job, Different Names: Matching Services

Most cloud services have a twin on the other two clouds. Use this table to translate a tutorial written for one cloud into the names on another.

Azure vs AWS vs Google Cloud Service Names
Common building blocks, matched by the job they do
The job Azure AWS Google Cloud
Virtual machines (computers you rent) Azure Virtual Machines Amazon EC2 Compute Engine
File and object storage Azure Blob Storage Amazon S3 Cloud Storage
Serverless functions (code that runs on demand) Azure Functions AWS Lambda Cloud Run functions
Containers without managing servers Azure Container Apps AWS Fargate Cloud Run
Kubernetes Azure Kubernetes Service (AKS) Amazon EKS Google Kubernetes Engine (GKE)
Relational databases (SQL) Azure SQL Database Amazon RDS Cloud SQL
NoSQL databases Azure Cosmos DB Amazon DynamoDB Firestore
AI platform for apps and agents Microsoft Foundry Amazon Bedrock Gemini Enterprise Agent Platform (formerly Vertex AI)
Sign-in and identity Microsoft Entra ID AWS IAM and IAM Identity Center Cloud Identity and IAM
Command-line tool Azure CLI AWS CLI gcloud CLI
Price estimates Azure Pricing Calculator AWS Pricing Calculator Google Cloud Pricing Calculator

A twin is rarely an exact copy. Limits, features and prices differ, so treat the table as a map, not a promise. For the full lists, Microsoft keeps an AWS to Azure services comparison and Google keeps a service comparison across all three clouds.

Working from Windows? All three command-line tools install on Windows 10 and Windows 11. Our Azure CLI for Windows guide walks through the Azure one step by step.

AI Platforms: Foundry vs Bedrock vs Gemini Enterprise Agent Platform

Each cloud now has one main place to build AI apps and agents. All three let you pick from models made by several companies, connect them to your own data, add safety filters, and pay per use. The differences are in which models sit at the center and which other services they plug into.

AZURE

Microsoft Foundry

Foundry, formerly Azure AI Foundry, hosts the OpenAI models Microsoft sells through Azure, plus models from Anthropic, Meta, Mistral, DeepSeek and others, with an agent service built in.

It fits best when your data already lives in Microsoft tools, such as SharePoint, Microsoft 365 or Azure databases. Our Microsoft Foundry guide covers it in depth.

Model provider tiles for OpenAI, Mistral AI, DeepSeek and Grok from the Models section of Microsoft's Foundry product page
AWS

Amazon Bedrock

Amazon describes Bedrock as the platform for building generative AI apps and agents at production scale, with access to hundreds of foundation models (large ready-made AI models) from several AI companies.

It fits best when your apps and data already run on AWS, since it works with AWS sign-in, storage and monitoring out of the box.

Amazon Bedrock product page describing it as the platform for building generative AI applications and agents at production scale
GOOGLE

Gemini Enterprise Agent Platform

Google renamed Vertex AI to Gemini Enterprise Agent Platform, shortened to Gemini platform on its own pages. Its Model Garden offers Google’s Gemini models, third-party models such as Anthropic’s Claude, and open models such as Gemma.

It fits best when you want Gemini models or work with BigQuery data. Older tutorials say Vertex AI; the ideas carry over.

Google Cloud Gemini platform page, formerly Vertex AI, with the headline Innovate, build, and deploy enterprise-ready agents and apps

Model lineups change every few months. A model that is only on one cloud today may land on another next quarter. Pick the platform for how it fits your data and team, and check the model catalog on the day you start.

Free Tiers and Trial Credits Compared

All three give new customers credit to spend, plus services that stay free up to monthly limits. Here is how the offers looked when we checked each vendor’s site in November 2026.

Free Offers for New Customers
From each vendor’s own free tier page. Amounts are in US dollars and can change, so check the live pages
Azure AWS Google Cloud
Starting credit $200 $100 at sign-up, plus up to $100 more for trying key services $300
How long the credit lasts 30 days Up to 6 months on the Free plan 90 days
Free for a limited time 20+ popular services free for 12 months (new customers) Free plan covers select services only Cloud SQL and Spanner have their own free trials
Always free (monthly limits) 65+ services 30+ services 20+ products
When the trial ends Move to pay-as-you-go or your services are disabled The Free plan account closes unless you upgrade to a Paid plan No automatic charges; you activate a paid account to keep going
AZURE

Azure free account

The Azure free account gives $200 of credit for 30 days, then popular services stay free for 12 months within monthly amounts. You add a card for identity checks, and Microsoft says it is not charged unless you move to pay-as-you-go.

Students can get Azure for Students with a school email and no credit card. Our Azure free account guide covers sign-up and the limits.

Azure free account card listing free monthly amounts of 20+ popular services for 12 months, 65+ always-free services, spending protection, and a 200 dollar credit to use within 30 days
AWS

AWS Free Tier

The AWS Free Tier now works in credits: $100 when you open the account, and up to $100 more as you try key services. The Free plan lasts up to 6 months or until the credits run out.

The Free plan cannot surprise you with a bill, but it closes on its own at the end unless you upgrade. The Paid plan unlocks every service and keeps the same credits.

AWS Free Tier page offering up to 200 dollars in credits for new customers, with Create a free account and Start building in console buttons
GOOGLE

Google Cloud free program

The Google Cloud free program gives new customers $300 of credit for 90 days. Google says you are not charged until you activate a full paid account.

Its free tier includes one small e2-micro virtual machine a month in select US regions, plus monthly free use of Cloud Storage, Cloud Run and BigQuery, which does not count against the $300.

Google Cloud free program page reading Free to sign up, Free monthly usage, with 300 dollars in free credit and products with a free tier

Free tiers have edges. A free virtual machine is small, often limited to certain regions, and storage or data sent out of the cloud can still cost money. Set a budget alert on day one, whichever cloud you pick.

How Pricing Works on Each Cloud

All three start with pay-as-you-go: you pay for what you use, by the second, hour, gigabyte or request, with no contract. Then each one offers discounts if you commit to steady use or accept servers that can be taken back. We do not list prices here, because they differ by region, size and day; use each cloud’s calculator instead.

Ways to Pay Less
The vendors’ own maximum discount claims versus pay-as-you-go. Real savings depend on region, size and usage
How you pay Azure AWS Google Cloud
Pay as you go Standard rates, no commitment On-Demand, no commitment On-demand, no commitment
Commit to a specific server type for 1 or 3 years Reservations, up to 72% off Reserved Instances, up to 72% off Resource-based committed use discounts, up to 55% off (70% for memory-optimized)
Commit to an hourly spend for 1 or 3 years Savings plan for compute, up to 65% off Savings Plans, up to 72% off Spend-based committed use discounts
Spare capacity that can be taken back Spot VMs, up to 90% off Spot Instances, up to 90% off Spot VMs, up to 91% off
Bring your own Microsoft licenses Azure Hybrid Benefit (see the next section) Limited; mostly on Dedicated Hosts Limited; mostly on sole-tenant nodes

Google also gives sustained use discounts automatically on some virtual machine types that run for a large part of the month, up to 30% off with no commitment. The other two need you to sign up for a reservation or plan to get a discount.

Try your own setup in the Azure Pricing Calculator, the AWS Pricing Calculator and the Google Cloud Pricing Calculator. Our Azure pricing guide explains how meters and bills work, and our guide to reducing Azure costs covers budgets, reservations and Advisor.

Data leaving the cloud costs money on all three. Sending data out to the internet or to another cloud (called egress) is billed per gigabyte, while sending data in is usually free. It is easy to forget when you compare server prices alone.

The Windows and Microsoft Licensing Angle

For Windows users and Windows shops, this is often the deciding section. Every cloud can run Windows Server, but they charge for it differently, and Microsoft’s own rules give Azure some advantages the other two cannot match.

AZURE

Azure Hybrid Benefit

Azure Hybrid Benefit lets you use Windows Server and SQL Server licenses you already own, with active Software Assurance or a qualifying subscription, so you pay only for the computer and not the license again.

Microsoft says it saves up to 80% on Windows Server and up to 85% on SQL Server versus pay-as-you-go rates, depending on region and size. Software Assurance comes with certain business licensing agreements, not with a retail copy of Windows.

Azure Hybrid Benefit page with the tagline Migrate to Azure, maximize your savings, and modernize your environment
LICENSES

Your licenses on AWS and Google

Microsoft names Amazon, Google and Alibaba as Listed Providers. Since October 2019, licenses bought without Software Assurance and mobility rights cannot go on their dedicated servers. AWS says Windows Server BYOL (bring your own license) needs Dedicated Hosts and licenses bought before October 1, 2019, so most people pay the license-included hourly price on AWS and Google Cloud.

UPDATES

Security updates for old Windows Server

Microsoft’s Extended Security Updates (extra years of security fixes after support ends) have been free for servers hosted in Azure and a paid subscription elsewhere. If you still run an older Windows Server, that can tip the math.

IDENTITY

Microsoft 365 and Entra ID sign-ins

If your company uses Microsoft 365, your work accounts already live in Microsoft Entra ID, the same identity system Azure uses. AWS and Google Cloud can sign people in with Entra ID too, but it is an extra setup step.

DESKTOPS

Windows desktops in the cloud

Azure Virtual Desktop, Microsoft’s service for running Windows desktops and apps in the cloud, runs on Azure. The other clouds offer their own virtual desktop services under their own terms.

Licensing rules are detailed and they change. If real money rides on it, check your agreement with your Microsoft licensing partner before you move servers. Microsoft’s migration tools can estimate Azure costs with and without Hybrid Benefit.

Which Cloud Should You Choose?

Find the card that sounds most like you. None of these is a rule, but each one is where we would start.

AZURE

You run a Windows or Microsoft 365 business

Start with Azure. Entra ID sign-ins, Hybrid Benefit for the licenses you already pay for, and free security updates for older servers all point the same way.

AWS

Your team already builds on AWS

Stay on AWS unless there is a clear reason to move. Skills, scripts and habits are worth real money, and moving clouds takes months.

GOOGLE CLOUD

You live in data, Kubernetes or Gemini

Look hard at Google Cloud. BigQuery, GKE and the Gemini models are its home turf, and its automatic discounts suit servers that run all month.

LEARNERS

You are learning for a job

Check the job listings where you want to work and learn the cloud they name most. The ideas transfer: once you know one cloud, the second one is mostly new names.

AI

You want to build an AI app or agent

Choose by where your data already sits: Foundry for Microsoft data, Bedrock for AWS, Gemini platform for Google. Then compare the models each one offers that week.

SMALL SITES

You just need a small website

You may not need any of the three. A simple web host is often cheaper and easier. Reach for a big cloud when you need its scale or its services.

Frequently Asked Questions

Which cloud is cheapest: Azure, AWS or Google Cloud?
None of them is always cheapest. Prices depend on the service, the region, the size and how you commit, and each cloud wins some comparisons. Price your real setup in each vendor’s pricing calculator. If you already own Windows Server or SQL Server licenses with Software Assurance, Azure Hybrid Benefit often makes Azure cheaper for those workloads.
Is Azure only for Windows?
No. Azure runs Linux virtual machines, containers, Kubernetes and open-source databases alongside Windows. Its advantage for Windows is licensing and Microsoft integration, not a lack of Linux support.
Which cloud should I learn first?
Learn the one that the jobs you want ask for, or the one your company uses. The core ideas, such as virtual machines, storage, networking, identity and billing, are the same on all three, so a second cloud is much faster to pick up than the first.
Can I use more than one cloud at the same time?
Yes, and many companies do. Tools such as Azure Arc let you manage Windows and Linux servers that run on another cloud provider from the Azure portal. Using two clouds does add work, since you have two bills, two sets of sign-ins and two sets of skills to keep up.
Can I move from AWS or Google Cloud to Azure later?
Yes. Azure Migrate can discover servers, estimate costs and move virtual machines from other clouds or your own datacenter. Plan for testing time, and remember that data sent out of your old cloud is usually billed per gigabyte.
Is my data equally safe on Azure, AWS and Google Cloud?
All three run large security teams and hold the major security certifications. All three also use a shared responsibility model: the provider secures the datacenters and hardware, and you are responsible for your accounts, settings and data. Most cloud breaches come from settings left open, so turn on multi-factor sign-in on any cloud you use.

Support and Community

More Windows software guides: What Is Microsoft Azure? · Azure Free Account · Azure Pricing Explained · Microsoft Foundry · Azure Migration Tools · Azure CLI for Windows


Discover more from Windows Mode

Subscribe to get the latest posts sent to your email.