AWS vs Azure vs Google: Which Cloud Is Best for Your Organization

Subscribe banner

Public cloud services are popular due to their high scalability, high availability, and numerous flexible options. The number of cloud providers is growing, but there are a few trusted cloud service providers that have been on the market for many years. The best-known cloud vendors are Amazon, Microsoft, and Google offering Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Services respectively. All of them are attractive and provide nice features for Infrastructure as a Service (IaaS), Software as a Service (SaaS), and Platform as a Service (PaaS). Choosing one of them can be difficult. AWS vs Azure vs Google cloud – which cloud service should you select? This blog post compares these three cloud-based platforms.

A Short History of Each Cloud Platform

Amazon is the pioneer in cloud services. The Amazon cloud platform is the oldest public cloud platform, dating back to 2006, and this company has been dominating the market ever since. AWS focuses on the public cloud rather than a hybrid or private cloud. Why use AWS? Read about the features below to decide whether AWS is suitable for you.

The Azure cloud platform has been on the market since 2010. Microsoft decided to complement their wide range of software and build a public cloud in their own data centers to attract more customers. Microsoft is now among the top three players for public cloud services. Microsoft demonstrated great progress with the Azure cloud platform.

What is GCP? Google Cloud Platform (GCP) was created in 2011 to provide Google cloud services. GCP is the youngest cloud platform and is growing quickly. GCP improves Google’s IaaS, PaaS, and SaaS. Google has a great infrastructure with data centers used for Google search services, YouTube, and Gmail. Google cloud services use the same infrastructure, and Google Cloud Platform has the highest growth rate in the cloud services market.

Virtual Machines

All cloud-based platforms provide the ability to run virtual machines (VMs), select different configurations for VMs, and select a VM class. Disk, CPU, memory, and input/output operations per second (IOPS) all depend on the selected VM class. Virtual machines and storage are the most used services in a cloud platform.

AWS. Virtual machines running in Amazon Web Services are called Elastic Compute Cloud (EC2) instances. You can select EC2 instances with preconfigured settings or configure virtual hardware settings manually. Amazon EC2 instances can run in different locations, such as data centers in different geographical regions. AWS has the highest variety of available data centers of all cloud platforms.

Azure. You can run Azure virtual machines in the Azure cloud. One of the main advantages of Azure virtual machines is that they use real processor cores. If, for example, you configure a VM to use one processor with four cores, AWS and Google Cloud Platforms create a VM processor with two cores and four threads (using hyperthreading). A VM in Azure has one processor with four real cores (without hyper-threading). This results in higher CPU performance for VMs with similar configurations running in Azure compared with other cloud platforms. VMs in Azure show excellent results in terms of performance.

Google Cloud uses Google Compute Engine to run virtual machines in Google Cloud Platform. Google does not offer wider VMs compared to AWS and Azure, but Google is more focused on containers and Kubernetes for running horizontally scalable applications with a microservice architecture.

The maximum VM computing configuration parameters for AWS vs Azure vs Google cloud platforms are compared in the table below.

AWS Azure Google Cloud
CPU 1.6 GHz – 3.3 GHz 2.7 GHz – 3.7 GHz 2.0 GHz – 4.0 GHz
Maximum vCPUs 128 128 224
Maximum Memory 244 208 448
Temporary Storage 48 TB 3 TB 4 TB
Maximum vGPUs 4 4 4

Container Support

All three cloud platforms support running containers, which are now extremely popular among developers of applications that use microservices.

Google played an important role in developing Kubernetes for container orchestration and, as a result, Google Cloud Platform has good support for Kubernetes and Docker containers. Google Cloud Run is used to develop and deploy containerized applications that must be highly scalable.

Amazon provides Amazon Elastic Container Registry, Amazon Elastic Container Service, and Amazon Elastic Kubernetes Service. Container services support Kubernetes, Docker containers, and Fargate services (Amazon EC2 Container Service).

Azure has two container services: Azure Kubernetes Service (AKS) and Azure Container Service (ACS). The Docker hub and Azure Container Registry are used to managing containers.

The table below lists each container service in AWS vs Azure vs Google Cloud.

Service AWS Azure Google Cloud
Docker container services Elastic Container Registry (ECR) Container Registry Container Registry
Managed container services EC2 Container Service (ECS)
Amazon Kubernetes Service
Azure Container Service (ACS) Google Kubernetes Engine
Serverless container services AWS Fargate Azure Container Instances (ACI) Google Cloud Run


Read more about Docker containers and Kubernetes in our blog posts.

Cloud Storage

Cloud storage is another important cloud platform service that is widely used by customers around the world. Cloud storage is often covered in AWS vs Azure vs Google Cloud comparisons. Each cloud platform provides different types of cloud storage.

Amazon cloud storage

Amazon S3 is an object-level storage service. All files and folders are stored as objects in Simple Storage Service (S3) buckets.

Amazon Elastic Block Storage (EBS) is a block-based storage service. EBS volumes are connected to Amazon EC2 instances to provide virtual disks for Amazon virtual machines.

Amazon Glacier is cold storage for rarely used data, for example, backups.

Elastic File System (EFS) is a scalable file system in the cloud for Linux that can be connected to EC2 instances running in the cloud and to on-site machines. NFSv4 is usually used to connect machines to EFS. General workloads and file sharing are popular EFS uses (configuring a file server, storing application data).

Storage Gateway is a special service configured in the cloud and on-site (on a virtual machine) to connect local machines with AWS cloud storage.

Azure Storage Platform

Azure Files is a universal storage service to share files with virtual machines running in Azure and local machines running on-site.

Azure Blobs is scalable storage for big data, including text data and binary data.

Azure Disks are block-level storage used as volumes for Azure VMs.

Azure Tables stores structured data for NoSQL databases (schemaless).

Azure Queues or Azure Queue Storage is a special storage type for a large number of messages that are used by applications to communicate between application components.

Google Cloud Storage

Persistent disks are block storage for virtual machines running in the Google Cloud (Google Cloud Compute Engine). Persistent disks are also used for Google Kubernetes Engine Service.

Object storage, with features such as versioning and access permissions, uses buckets to store objects.

Filestore is network file storage used to store, share, and access data over a network.

The cloud storage options with AWS vs Azure vs Google Cloud are listed in the table below.

Service AWS Azure Google Cloud
Block Storage Elastic Block Storage (EBS) Azure Disk Storage Google Persistent Disks
Object Storage Simple Storage Service (S3) Azure Blob Storage Google Cloud Storage
File Storage Elastic File System (EFS) Azure Files Google Cloud Files
Archive Storage S3 Glacier Deep Archive
S3 Infrequent Access
Azure Archive Storage
Azure Cool Blob Storage
Google Cloud Storage Nearline, Coldline, and Archive
Bulk Data Transport AWS Snow Family

AWS Import/Export Service

Azure Data Box

Azure Import/Export Service

Storage Transfer Service

Network Services

Network services allow you to create virtual networks (and connect virtual machines running in the cloud to these networks), configure routing and access within your onsite environment or between cloud environments and provide load balancing for networks.

All three providers have similar network services and capabilities. AWS, Azure, and Google Cloud provide network redundancy for their services. In order to have lower network latency, select a data center region that is geographically closest to your organization’s physical location.

AWS core networking services use an internal architecture called a virtual private cloud (VPC), which is a completely isolated logical network. Google uses the Andromeda architecture for networking. This is Google’s network virtualization stack. The internal architecture of Azure networks is closer to the traditional network architecture of data centers and private networks. Azure Virtual Networking (VNet) is Microsoft’s core cloud network product.

Service AWS Azure Google Cloud
Direct Connection AWS Direct Connect Azure ExpressRoute Google Cloud Interconnect
Global Content Delivery Networks (CDNs) Amazon CloudFront Azure CDN Google CDN
DNS Amazon Route 53 Azure DNS
Traffic Manager
Google Cloud DNS
Virtual Private Cloud (VPC)  Network VPC Virtual Networks (VNet) Google VPC
Load balancing Elastic Load Balancing (ELB) Application Gateway

Azure Load Balancer

Cloud Load Balancer


When you connect your onsite infrastructure (for example, VMware vSphere) with the public cloud infrastructure and configure network connections between them, you get a hybrid cloud deployment model. Microsoft Azure provides a wide range of hybrid options for Microsoft clients.


A firewall allows you to configure access only to what you need and only from allowed sources. All three cloud platforms provide a managed firewall to configure secure network access to virtual machines and services on them. In the AWS vs Azure vs Google Cloud comparison, firewalls have many similar options.


AWS provides the AWS Network Firewall, a managed service that can be managed in AWS Firewall Manager. The AWS Firewall is divided into two categories, Network Firewall and Web Application Firewall. The network firewall is used to filter network traffic for the appropriate network protocols such as IP addresses, ports, etc. The network firewall includes packet filtering, a virtual private network (VPN), deep packet inspection, website filtering, and DNS reputation filtering.

The AWS Web Application Firewall provides application security and traffic filtering. Application security is used to protect web applications from attacks such as distributed denial of service (DDoS) attacks, zero-day attacks, data leaks, etc. Traffic filtering is based on HTTP headers, IP addresses, keywords, and URI strings. Users can use third-party firewalls available in the AWS Marketplace in addition to AWS firewalls.


Azure firewall services include Azure Firewall Premium, Azure Application Gateway, and Azure Web Application Firewall. Each firewall service is intended for specialized purposes. The Azure firewall offers Network Address Translation (NAT) and filtering for IP addresses, Transport Control Protocol (TCP) and User Datagram Protocol (UDP) ports, and HTTPS traffic. In addition, Azure Firewall Premium includes Intrusion Detection and Protection System (IDPS) and TLS inspection. Azure Application Gateway acts as a load balancer for HTTPS traffic and a reverse proxy that can encrypt and decrypt Secure Socket Layer (SSL) traffic. Azure Application Gateway supports web traffic inspection and attack detection at the HTTP level. Azure Application Gateway has an addition called Azure Web Application Firewall (WAF) that is used to inspect HTTP requests and prevent malicious web attacks, Cross-Site Scripting (CSS), and SQL injection. Azure firewall services complement each other. If we look at the Azure firewall as a network firewall and a web application firewall, we can categorize the protection types for each firewall as follows:

  • Azure Network firewall includes Entry Point Protection, VPN support, Software-Defined Wide Area Network (SD-WAN) capabilities, virtual WAN support, and identity and access management.
  • Azure web application firewall includes traffic filtering, script protection, secure delivery, customized rule sets, API protection, and security.

Google Cloud Platform

Google Cloud also provides a firewall. You can configure firewall rules for ingress/egress traffic and secure network access to virtual machines running in Google Cloud Platform. Firewall options are part of the VPC network configuration. Firewall rules for Google VPC work similarly to AWS security groups.


A firewall helps improve network security onsite and in the cloud. However, there are additional features for cloud-based platforms that improve security. In this cloud providers comparison, all three cloud platforms provide an excellent level of security. An encrypted connection is used to access cloud services on each cloud platform. However, customers may need to check and edit the security configuration to meet their security requirements.


AWS uses security isolation as the default principle when you create an account, a virtual machine, or other objects to protect cloud resources against unauthorized access. The security policy is strict by default. Some security tools can be supported in particular regions and not work in other regions.


One of the most popular security features of the Azure cloud platform is the Azure Active Directory. Active Directory is the centralized authentication service developed by Microsoft for secure authentication of Windows machines and supported software. Azure Active Directory allows you to integrate the onsite Active Directory of your local Active Directory domain with Azure Active Directory in the cloud. You can configure Active Directory federation services.

If you create an object in the cloud, security configuration by default is not as strict as in AWS. AWS and Google Cloud use the default Deny policy in access configuration, while Azure uses the Allow policy. For example, if you create a new virtual network and a new VM in Azure, all protocols and ports are opened by default.

Azure Activity Logs and Azure Security Center provide many advantages compared to AWS. You don’t need to build Lambda functions manually to move events between regions when you use Azure with the Activity Logs feature. Documentation in AWS is more detailed than in Azure.

Google Cloud Platform

Google Cloud Platform is more centralized, similar to Azure. When Google launched Google cloud services, all services were planned to interact well with other services and were launched at once (in AWS, services were added periodically and consequently). Projects in your account are isolated from each other by default. The Cloud Security Command Center in Google Cloud is equivalent to the Azure Security Center. The approximate level of security in Google Cloud is somewhere in between AWS and Azure security.

AWS Security Hub, Azure Security Center, and Cloud Security Command Center in Google Cloud are the security management tools for each cloud platform.


All three vendors provide the database as a service (DBaaS) option for customers. With DBaaS, customers can work with databases without needing to manage the infrastructure to run databases. Relational databases and NoSQL databases are supported.


AWS provides the broadest range of database options. Solutions work with high performance, innovations are implemented on time, and traditional database technologies are available. You can select AWS database services if you already use other AWS services, you expect a high level of performance and reliability, or you need the broadest set of options.


Azure provides great support for migration, including migration assessment, automation, and optimization. There are flexible deployment options, licensing options, and hybrid deployment is available (for those who respect security and privacy). You might select Azure databases if you already use Microsoft software in your environments (including a Microsoft-based hybrid environment), if you need to migrate a database to the cloud, and if privacy is a particular concern.

Google Cloud Platform

Database services in Google Cloud are the most user-friendly and provide the best performance for workloads. Google offers excellent capabilities to use databases with containers in Google Cloud. You may prefer Google databases if you need to attach a database to containers (for the microservice architecture), if you need high performance and a user-friendly solution.

AWS Azure Google Cloud
Relational database Amazon RDS MS SQL Database Google Cloud SQL
NoSQL Key-Value Amazon DynamoDB Table Storage Google Cloud Bigtable

Google Cloud Datastore

NoSqL Key-Index Amazon SimpleDB Azure Cosmos DB Google Cloud Datastore


You can read more about databases on our blog:

How to configure MS SQL Server replication?

How to install Oracle on Ubuntu?

Regions and Availability Zones

Each cloud provider covers these major areas with data centers: Europe, North America, Southeast Asia, East Asia, and China. Data centers are distributed within units called regions and availability zones.

A region is a set of data centers built in a particular (separate) geographical area. A region is an area where data centers physically exist. Data centers are connected with each other by low-latency networks (the latency-defined perimeter). Regions are the largest cloud provider units that contain availability zones. One region is completely independent of other regions.

An availability zone is a unique physical location within a region. Availability zones are isolated from each other within a region and are connected with each other by hi-speed redundant networks. If one availability zone fails within the region, other functioning availability zones provide the needed services to customers. An availability zone consists of one or more data centers.

AWS. Amazon provides over 80 availability zones in 25 geographic regions.

Azure. There are more than 60 regions in Azure with at least 3 availability zones per region. The Azure cloud platform has more than 160 physical data centers in 140 countries.

Google Cloud Platform. There are 24 regions and 73 availability zones.

AWS Azure Google Cloud
Regions 25 60+ 24
Availability Zones 80 180+ (at least 3 per region) 73
Point of Presence (POP) 230 130 144*
Countries 245 140 200
* Network Edge Locations


Periodically, providers add availability zones and data centers in different countries. See the detailed updated list of cities and other data center locations of data centers on each cloud provider’s website. The map of data center locations can help you select data centers in the needed location.


Price is an important factor that impacts the choice of the cloud platform. Knowing the price helps you estimate how much you need to spend on cloud services. It is difficult to compare AWS vs Azure vs Google pricing because prices change with time. The main costs are usually for computing services such as virtual machines. The price depends on the region where a data center is located, the CPU configuration of a VM, the amount of memory, disk space, and disk type (SSD or HDD).

Billing is provided on a per-hour and per-second basis. If you pay for a 1-year commitment with one transaction (or more, for example, three years), you can get a discount. In this case, you usually should select a reserved instance of the needed type.

For a correct cloud providers comparison in terms of pricing, we should select a similar region for all three providers and a similar VM configuration. AWS, Azure, and Google provide pre-configured virtual machines (you have to select a configuration preset). In the table below, you can see four types of virtual machines selected for this AWS vs Azure vs Google comparison. Some Google VMs have more memory and CPUs because there is no 100% identical configuration in the appropriate class of Google VMs in the first example. The most identical configuration for a Google VM is selected in this case.

Example 1. See the first table with types of virtual machines and the second table with the price for them.

Table 1: Types of instances (virtual machines).

Instance Type AWS Instances AWS RAM (GB) Azure VMs Azure RAM (GB) Google VMs Google RAM (GB)
General Purpose m6g.xlarge 16 B4MS 16 e2-standard-4 16
Memory-Optimized r6g.xlarge 32 E4a v4 32 m1-ultramem-40 961
Compute Optimized c6g.xlarge 8 F4s v2 8 c2-standard-4 16
Accelerated Computing p2.xlarge 61 NC4as T4 v3 28 a2-highcpu-1g 85


Let’s check the per-hour price at the time of writing (November 2021) for the selected configuration of virtual machines.

Table 2: On-demand pricing (USD).

Instance Type AWS Azure Google AWS pricing (per hour) Azure pricing (per hour) Google pricing (per hour)
General-purpose m6g.xlarge B4MS e2-standard-4 0.154 0.166 0.156
Memory optimized r6g.xlarge E4a v4 m1-ultramem-40 0.202 0.252 6.303
Compute optimized c6g.xlarge F4s v2 c2-standard-4 0.136 0.169 0.235
Accelerated computing p2.xlarge NC4as T4 v3 a2-highcpu-1g 0.90 0.526 3.839


The price for VMs in AWS and Google Cloud is similar for general-purpose VMs and memory-optimized VMs. The price difference between the Azure cloud platform and AWS cloud service for compute-optimized VMs is negligible. This is only one example, and if you select a 1-year commitment, a different provider can have the cheapest price for an instance type. Moreover, there are different prices for containers, storage, database services, and other types of cloud computing.

Example 2. Let’s select the smallest virtual machine and the largest virtual machine for each platform with identical parameters and compare the per-month price.

Table 1: Configuration of virtual machines.

VM type AWS CPU AWS RAM Azure CPU Azure RAM Google CPU Google RAM
Smallest 2 CPUs 8 GB 2 CPUs 8 GB 2 CPUs 8 GB
Largest 128 CPUs 3.84 TB 128 CPUs 3.89 TB 160 CPUs 3.75 TB


Table 2: Price (USD) for the selected VMs.

VM Type AWS Azure Google Cloud
Smallest $69/month $70/month $52/month
Largest $3.97/hour $6.79/hour $5.32/hour


In this example, the price for the smallest instance in AWS and Azure is almost the same, but the price in Google Cloud Platform is significantly less. As for the largest VM instance, AWS offers the lowest price and Azure offers the highest price. Remember that VMs in Azure use real CPU cores, unlike VMs in AWS and Google cloud where logical cores (hyper threading cores) are used. Real cores provide higher performance.

As you can see from these examples, the best price can be different in different scenarios.

Note: Pricing can change with time. To get the latest price, please check pricing information on the AWS, Azure, and Google Cloud websites.

Storage costs

Object storage. You should keep in mind the main differences between object storage pricing in AWS and Google Cloud and the approach used to determine the price. In Google Cloud, you must pay for operations performed on object storage and network egress. Data access modeling is recommended before calculating costs. In Google Cloud Platform, you have instant access to all infrequent storage tiers. In AWS, access time to Amazon archive storage ranges from minutes to hours.

Block storage. If we compare AWS with Google Cloud, we will find some differences. Google Cloud provides high availability within the entire region across availability zones and across multiple regions. AWS provides redundancy only inside the same availability zone. AWS has an extra charge for provisioned IOPS that allows EBS volumes to burst over their usual data transmission rates. There is no IOPS limit in Google Cloud for Google block storage and you don’t pay for extra IOPS.


In general, AWS pricing is complicated, and it’s difficult to understand the cost structure, especially for new customers. To get a discount, AWS requires prepayment for reserved instances that are for long-term usage.

If a VM is stopped, you are charged only for storage space used by the EBS volumes.

A 12-month free trial is provided for new AWS users.


Microsoft software is popular among customers and is widely used by organizations. This is one of the reasons for Microsoft’s success as a cloud provider. Discounts are provided for existing Microsoft customers who sign into Azure and use AWS cloud services. You should familiarize yourself with Microsoft software licensing options when starting to use Azure.

There is a 5% discount for a 12-month prepayment.

Power off VMs correctly, without preserving the IP address obtained by a VM. A VM must be deallocated to avoid charges if the VM is not running in Azure.

The free trial period for new Azure customers is 12 months and includes $200 that can be spent within the first 30 days after registration and starting the trial. More than 25 Microsoft products in Azure are offered for the trial period.

Google Cloud Platform

Google Cloud offers a user-friendly pricing structure. There are discounts for long-running workloads without an up-front commitment.

When you stop a VM, you are not charged for VM computing resources such as CPU, GPU, or memory, but you are charged for resources attached to the VM such as persistent disks and static IP addresses.

Google provides a $300 credit for 90 days for new users who are starting the free trial period. More than 20 products from Google cloud services are offered for trial users.

Cost optimization

There are cost optimization tools that can help you select the optimal configuration of services on a selected cloud-based platform.

AWS: AWS Cost Explorer, AWS Trusted Advisor, AWS Budgets

Azure: Azure Advisor

Google Cloud Platform: Cost Management

A price breakdown is complicated in a cloud provider's comparison because different pricing models are used by each cloud platform. Please use the AWS price calculator, Azure price calculator, and Google Cloud price calculator to get the exact price for the needed configuration and to compare prices. Using the calculator is the best way to estimate monthly expenses for the cloud services required.

Data Analytics and Machine Learning Services

All three vendors offer data analytics services, machine learning (ML), and artificial intelligence (AI). These types of cloud computing are widely used nowadays for data analysis, science, research work, automation, etc. ML usually contains data pre-processing, model training, model evaluation, event prediction, image recognition, etc. A highly scalable computing cloud is a suitable platform to run these tasks. Amazon cloud platform, Azure cloud platform, and Google Cloud Platform provide machine learning as a service (MLaaS).

The older ML service in AWS is called Amazon Machine Learning and the newer one is SageMaker. Amazon Machine Learning is primarily used for predictive analytics and SageMaker is preferred by data scientists. Both Amazon and Azure offer integration with Jupiter that allows you to write code in ML Studio. One of the top ML services provided by Google is Vision AI (powered by Auto ML).

AWS AI/ML services (12):

  • SageMaker
  • Machine Learning
  • Comprehend
  • Lex
  • Polly
  • Rekognition
  • Translate
  • Transcribe
  • DeepLens
  • Deep Learning AMIs
  • Apache MXNet on AWS
  • TensorFlow on AWS

Microsoft Azure AI Platform (3 services):

  • Machine Learning
  • Azure Bot Service
  • Cognitive Services

Google AI Platform (9 services):

  • Cloud Machine Learning Engine
  • Dialogflow Enterprise Edition
  • Cloud Natural Language
  • Cloud Speech API
  • Cloud Translation API
  • Cloud Video Intelligence
  • Cloud Job Discovery (Private Beta)

See the list of available ML/AI features in the comparison of AWS vs Azure vs Google Cloud in the table below.

Cloud providers comparison – machine learning.

Amazon ML and SageMaker Microsoft Azure AI Platform Google AI Platform
Classification + + +
Regression + + +
Clustering + + +
Anomaly detection + + -
Recommendation + + +
Ranking + + -
Data Labeling + + +
MLOps pipeline support + + +
Built-in algorithms + + +
Supported frameworks TensorFlow, MXNet, Keras, Gluon, PyTorch, Caffe2, Chainer, Torch TensorFlow, scikit-learn, PyTorch, Microsoft Cognitive Toolkit, Spark ML TensorFlow, scikit-learn, XGBoost, Keras

Machine learning APIs

Besides excellent and powerful cloud platforms with services that are ready to use, you can use high-level APIs to work with your custom applications. You can use these services with ready-trained models, feed your data (input), and get results (output).

In this AWS vs Azure vs Google comparison, APIs are divided into three groups:

  • Text translation, recognition, and analysis
  • Video and image recognition and analysis of these content types
  • Other uncategorized services

A comparison of speech and text processing APIs is in the table below.

AWS Azure Google Cloud
Speech Recognition (Speech into text) + + +
Text into speech conversion + + +
Entities Extraction + + +
Key Phrase Extraction + +
Language Recognition 100+ languages 120 languages 120+ languages
Topics Extraction + + +
Spell Check - + -
Auto completion - + -
Voice verification + + -
Intention Analysis + + +
Metadata Extraction - - -
Relations Analysis - + -
Sentiment Analysis + + +
Personality Analysis - - -
Syntax Analysis - + +
Tagging Parts of Speech - + +
Filtering inappropriate content - + +
Low-quality audio handling + + +
Translation 6 languages 60+ languages 100+ languages
Chatbot Toolset + + +


Microsoft offers the widest set of features in the cloud providers comparison for machine learning APIs.

A comparison of versatile APIs for image analysis is displayed in the next table.

AWS Azure Google Cloud
Object Detection + + +
Scene Detection + + +
Face Detection + + +
Face Recognition + + -
Person Face Identification + + +
Facial Analysis + + +
Inappropriate content detection + + +
Celebrity Recognition + + +
Text Recognition + + +
Written Text Recognition + + +
Search for similar images on the Web - - +
Logo Detection - - +
Landmark Detection - + +
Food Recognition + + -
Dominant Colors Detection - + +


Google Cloud Platform offers the most versatile toolkit for image analysis.

Comparison of video analysis APIs

The process of video analysis has similarities with the process of image analysis, but in the AWS vs Azure vs Google Cloud comparison of video analysis APIs, the cloud provider ranking is different. In contrast to image processing support, Google doesn’t provide a rich set of APIs for video analysis and many features are still in the development or beta version phase. Amazon and Microsoft provide a wider set of video analysis APIs and related features.

AWS Azure Google Cloud
Object detection + + +
Scene detection + + +
Activity detection + - -
Facial Recognition + + -
Facial and Sentiment Analysis + + -
Inappropriate Content Detection + + +
Celebrity Recognition + + -
Text Recognition + + -
Person Tracking on Videos + + -
Audio Transcription - + +
Speaker Indexing - + -
Keyframe Extraction - + -
Video Translation - 9 languages -
Keywords extraction - + -
Brand Recognition - + -
Annotation - + -
Dominant Colors Detection - - -
Real-Time analysis + - -


In the video APIs comparison of AWS vs Azure vs Google Cloud Platform, Microsoft gets the highest score and is the leader. However, AWS offers the most efficient APIs for video analysis of streaming videos.

This AWS vs Azure vs Google Cloud comparison in the ML/AI category is a brief overview of features. If you dig deeper, you can explore more interesting information and find more details to compare. However, a detailed comparison of cloud providers for AI/ML types of cloud computing is beyond the scope of this blog post.


The AWS vs Azure vs Google Cloud comparison is complex because each cloud platform offers a wide set of features. When you compare Amazon cloud platform, Azure cloud platform, and Google cloud services, focus on the services that you need, first of all. This blog post has covered a comparison of these three cloud platforms as an overview of the most popular service categories.

AWS is the most vendor-locked provider aimed at making you use the Amazon cloud platform only. In contrast, Google provides a flexible and liberal policy for customers. Microsoft wants to mix the advantages of AWS and Google Cloud Platform and integrate Azure with other solutions and providers. Microsoft offers the best hybrid cloud options that allow you to use the Azure cloud with other clouds and with onsite servers in your local data center. One more advantage of Microsoft and Google is that they provide online office applications such as Microsoft 365 and G-Suite in addition to Azure and Google Cloud Platform.

All cloud solutions explained in this cloud platform comparison are reliable and can be used to store backups. However, you may need to back up data stored in the cloud to somewhere else. In this case, you should use a reliable backup solution that can protect data by performing backup in both directions. NAKIVO Backup & Replication is the universal data protection solution that supports backing up Amazon EC2 instances to a backup repository deployed onsite or in the cloud. You can back up data from onsite servers including Windows servers, Linux servers, VMware VMs, Hyper-V VMs, and other data to Amazon S3 buckets with NAKIVO Backup & Replication. Using Amazon S3 object lock for backups made with the solution helps you protect data against ransomware.

Download the free edition of NAKIVO Backup & Replication and try the professional backup solution to back up data.

Let’s Stay in Touch

Subscribe today to our monthly newsletter
so you never miss out on our offers, news and discounts.