The moment you start searching for how to learn cloud computing, you get hit with names like AWS, Azure, and GCP, plus unfamiliar terms like EC2, regions, and IAM, all at once. It is completely normal to wonder where to begin, whether a free account will quietly run up a bill, and which cloud certification you should take first. This is a step-by-step starter guide for cloud beginners that covers only what you actually need to know in your first two months.
Based on each provider's official pages as of November 2026, it walks through the core ideas of cloud computing, the differences between AWS, Azure, and GCP with a service name cheat sheet, a comparison of the three free tiers and the safety settings to turn on right after you sign up, three first hands-on labs, a comparison of beginner certifications, an 8-week study roadmap, and answers to common questions. If you are still torn about which cloud to choose, read AWS vs. Azure vs. GCP: Which Cloud Certification Should You Start With? as well.
---
Who This Guide Is For
| If you are... | What you will get from this guide | |------|------| | Completely new to cloud, with no technical background | Cloud concepts explained through everyday analogies, plus the terms you really need | | Experienced in IT or development, but new to cloud | A service name cheat sheet for all three providers and labs you can try today | | A job seeker or career changer choosing a first certification | A side-by-side look at four beginner certifications and which one fits your situation | | Keen to try cloud for free but worried about the bill | A comparison of the three free tiers and the settings that prevent surprise charges |
Short on time? Jump to the Key Takeaways table at the bottom, then come back to the sections you need.
---
What Is Cloud Computing? A 5-Minute Explanation for Beginners
Cloud computing means renting IT resources such as servers, storage, and databases over the internet, using only what you need and paying only for what you use, instead of buying and installing the hardware yourself. It works a lot like electricity: you do not keep a generator at home, you use power from the utility company and pay for what you consume. The three biggest "utility companies" of the cloud are AWS (Amazon Web Services), Microsoft Azure, and Google Cloud (GCP).
Companies move to the cloud for four main reasons.
On demand: getting a new server used to take weeks of ordering and installing; in the cloud it takes minutes. Pay as you go: no expensive upfront hardware, you pay only for the time and amount you use. Easy to scale: add servers when traffic spikes and remove them when things are quiet. Global reach: serve users in other countries from a data center close to them.
IaaS vs PaaS vs SaaS: The Housing Analogy
Cloud services fall into three groups depending on how much the provider manages for you. Comparing them to ways of finding a place to live makes the difference easy to remember.
| Model | Analogy | Provider manages | You manage | Examples | |------|------|------|------|------| | IaaS (infrastructure) | Renting an empty house | Building, power, water (servers, network) | Furniture and belongings (operating system, software, data) | Amazon EC2, Azure Virtual Machines, Compute Engine | | PaaS (platform) | A fully furnished apartment | Everything from the building to the furniture (OS, runtime) | Your own stuff (code, data) | AWS Elastic Beanstalk, Azure App Service, App Engine | | SaaS (software) | A hotel | Almost everything | Your account and your data | Microsoft 365, Google Workspace, Zoom |
Beginner tip: questions that ask you to tell IaaS, PaaS, and SaaS apart show up on all three providers' entry-level exams. Ask yourself "who manages the operating system?" and you will rarely get them wrong.
Regions and Availability Zones: Where the Data Centers Are
A region is a geographic area where a provider clusters its data centers. AWS, Azure, and Google Cloud all operate regions on every continent except Antarctica, so you can pick one close to you or to your users. Most regions contain several availability zones, which are groups of data centers with separate power and networking, so if one location fails, your service can keep running from another.
The Shared Responsibility Model: The Cloud Doesn't Do Everything for You
The cloud provider is responsible for security "of the cloud": the data center buildings, server hardware, and network. You are responsible for security "in the cloud": the data you upload, your account passwords, and who has access to what. This split is called the shared responsibility model. It appears on all three entry-level exams, and in real life many account takeovers and data leaks start with a configuration mistake on the customer's side. You can read more in Shared Responsibility Model.
---
AWS vs Azure vs GCP: The Big Three at a Glance
All three offer similar services, but they started in different places and have different strengths. By market share, AWS is first, Microsoft Azure second, and Google Cloud third.
| Item | AWS | Microsoft Azure | Google Cloud (GCP) | |------|------|------|------| | Origins | 2006, grew out of Amazon's experience running its own infrastructure | Generally available in 2010, closely tied to Microsoft's business software | Started with App Engine in 2008, built on the infrastructure that runs Google's own services | | Strengths | The most services and customers, plus the largest pool of learning resources and community | Integration with Microsoft 365, Windows Server, and Entra ID | Data analytics (BigQuery), AI (Gemini), Kubernetes (GKE) | | Common users | Startups, internet services, gaming companies | Large enterprises that run Microsoft products, finance, manufacturing, government | Data and AI focused companies, engineering teams that run Kubernetes | | Beginner experience | Feels huge at first because of the number of services, but has the most study material | Familiar if you have used Microsoft products | Often described as simpler to navigate, with fewer third-party resources | | Flagship beginner certification | Cloud Practitioner (CLF-C02) | Azure Fundamentals (AZ-900) | Cloud Digital Leader |
Beginner tip: the core concepts are almost identical across all three. Once you truly learn one, picking up the other two is mostly a matter of finding "the same service under a different name." That is why it pays to focus on just one at the start.
---
Same Job, Different Names: AWS, Azure, and GCP Services Compared
Service names are what confuse beginners the most. The table below lines up each provider's main service for the same job. They are not identical feature for feature, but this is more than enough to find your way when you start out.
| Job | In plain words | AWS | Azure | Google Cloud | |------|------|------|------|------| | Virtual machines | A computer you rent over the internet | Amazon EC2 | Azure Virtual Machines | Compute Engine | | Serverless functions | Run code without managing servers | AWS Lambda | Azure Functions | Cloud Run functions | | Kubernetes | Run containerized apps at scale | Amazon EKS | Azure Kubernetes Service (AKS) | Google Kubernetes Engine (GKE) | | Object storage | A virtually unlimited warehouse for files and photos | Amazon S3 | Azure Blob Storage | Cloud Storage | | Relational database | Store data in tables | Amazon RDS | Azure SQL Database | Cloud SQL | | NoSQL database | Flexible, high-volume data | Amazon DynamoDB | Azure Cosmos DB | Firestore | | Data warehouse | Storage built for analyzing large datasets | Amazon Redshift | Microsoft Fabric · Azure Synapse Analytics | BigQuery | | Virtual network | Your own private network inside the cloud | Amazon VPC | Azure Virtual Network | VPC | | Content delivery (CDN) | Deliver content quickly from near your users | Amazon CloudFront | Azure Front Door | Cloud CDN | | Identity and access | Control who can do what | AWS IAM | Microsoft Entra ID · Azure RBAC | Cloud IAM | | Monitoring | Check health and logs | Amazon CloudWatch | Azure Monitor | Cloud Monitoring · Cloud Logging | | Generative AI platform | Pick an AI model and connect it to your app | Amazon Bedrock | Microsoft Foundry (formerly Azure AI Foundry) | Gemini Enterprise Agent Platform (formerly Vertex AI) | | Browser terminal | Run commands without installing anything | AWS CloudShell | Azure Cloud Shell | Cloud Shell |
Beginner tip: each certification exam covers a single provider, so learn the official names for the provider you are studying. Many services have been renamed recently, such as Vertex AI becoming Gemini Enterprise Agent Platform and Azure AI Foundry becoming Microsoft Foundry, so double-check any name you saw in older material against the official documentation.
To dig deeper into the core services, start with AWS EC2 Essentials, Amazon S3 Essentials, and Azure Virtual Machines Explained.
---
Start for Free: AWS Free Tier vs Azure Free Account vs Google Cloud Free Trial
All three providers give new customers free credits plus a set of always-free services. Here is how they compare according to the official pages as of November 2026. The terms change often, so check the official page once more before you sign up.
| Item | AWS Free Tier (Free plan) | Azure free account | Google Cloud Free Trial | |------|------|------|------| | New-customer credits | 100 USD at sign-up, plus up to 100 USD more for trying key services (up to 200 USD) | 200 USD | 300 USD | | How long | Up to 6 months | 30 days | 90 days | | Always free | 30+ services within monthly usage limits | 65+ always-free services, plus 20+ popular services free for 12 months | 20+ products within monthly usage limits | | Automatic charges | None until you choose to upgrade to the Paid plan | None until you choose to move to pay-as-you-go | None until you choose to upgrade to a paid billing account | | When it ends | The accou