Cloud Services Cross-Reference Guide: AWS, Azure, OCI & Google Cloud

This guide maps equivalent services across Amazon Web Services (AWS), Microsoft Azure, Oracle Cloud Infrastructure (OCI), and Google Cloud Platform (GCP). It is organized into twelve functional domains covering every major service category. The guide is intended for cloud architects, engineers, and developers who need to evaluate services across providers, plan multi-cloud deployments, or port workloads between clouds.

Each cloud provider uses different names for equivalent capabilities. This guide cuts through that terminology barrier. The master cross-reference table below provides a fast-scan view of service equivalencies, and the domain-specific guides linked in Section 4 provide detailed feature comparisons, differentiator analysis, and per-section tables. This index covers all four providers with equal depth; it is not a recommendation document.

Cloud providers evolve rapidly. Several services documented here reflect significant status changes as of early 2026: AWS Snowball Edge is closed to new customers, OCI Data Transfer reached end-of-life, GCP Cloud IoT Core was retired in 2023, AWS CodeCommit returned to general availability in late 2025, and GCP Deployment Manager is reaching end of support in March 2026. Domain guides note these current-state details.

Provider Overview

Master Cross-Reference Table

The following table covers the primary service equivalent in each functional category. Use the domain links in Section 4 for feature-level detail on each row.

Domain Guides

Each domain below provides full per-section tables, feature-level differentiators, and current-state notes on service changes and deprecations.

Key Differentiators

What each cloud does uniquely well, based on the 12 domain guides.

AWS

AWS leads in service breadth and ecosystem depth. With over 200 services and the widest range of compute instance families (600+ EC2 types), AWS provides a matching managed service for almost every workload pattern. Amazon Bedrock offers the largest foundation model catalog (100+ models from the most third-party providers) for generative AI applications. The AWS analytics ecosystem around Apache Iceberg — spanning Redshift, EMR, Athena, and Glue with unified Lake Formation governance — is the most mature open lakehouse implementation on any single cloud. AWS Lambda@Edge enables FaaS execution at CloudFront PoPs globally, with no equivalent from the other providers. SageMaker HyperPod with Trainium2 provides a vertically integrated ML training stack with automated fault recovery. AWS Systems Manager is the most comprehensive native operational automation platform, bundling patch management, run command, session management, configuration enforcement, and 300+ pre-built runbooks in a single service.

Azure

Azure provides the deepest enterprise integration. Microsoft Entra ID is the industry-dominant cloud directory, with native integration across Microsoft 365, Dynamics 365, Intune, Power Platform, and thousands of SaaS applications — making Azure the natural choice for Microsoft-centric organizations. Azure OpenAI Service is the exclusive enterprise deployment path for OpenAI's GPT-4o, o3, and DALL-E 3 models, under Azure's compliance, data residency, and private networking guarantees. Azure Arc is the only native tool from any cloud provider that extends a cloud's governance, monitoring, and policy plane to resources running in competing clouds or on-premises. Microsoft Fabric consolidates warehousing, BI (Power BI), data engineering, and real-time analytics into a unified SaaS platform with a single OneLake storage layer. Azure Migrate is the most integrated server and database migration hub, covering VMs, databases, web apps, and VDI in one assessment and orchestration surface. Microsoft Sentinel is the most feature-complete cloud-native SIEM/SOAR platform.

OCI

OCI differentiates on price-performance and Oracle workload depth. OCI's egress pricing is approximately one-tenth the rate of comparable AWS volumes (10 TB/month free, then a flat low rate), a structural advantage for data-intensive applications. OCI FastConnect charges only a per-port-hour fee with no data transfer fees, significantly undercutting AWS Direct Connect and Azure ExpressRoute for high-volume private connectivity. OCI NAT Gateways, Private Endpoints, and the DRG transit hub are provided at no additional charge — all equivalent services on AWS, Azure, and GCP carry per-hour and per-GB fees. OCI Flexible Shapes allow independent OCPU and memory selection at instance launch, which no other cloud offers for standard VMs. OCI Autonomous Database is uniquely self-patching, self-tuning, and self-securing with a 99.995% SLA covering both planned and unplanned downtime — no other cloud database offers this. OCI leads in raw GPU cluster scale: up to 131,072 GPUs in a single interconnected cluster with bare metal access and RDMA networking. Oracle Integration Cloud is the most fully featured enterprise iPaaS of the four providers, with native B2B/EDI and deep Oracle ERP connectors.

GCP

GCP leads in data analytics and AI/ML. BigQuery is the only cloud data warehouse that is architecturally serverless with no clusters, no node provisioning, and per-query pricing — simplifying data warehouse operations at any scale. GCP's Dataflow (Apache Beam) is the only streaming engine where the same pipeline code runs both batch and streaming workloads without modification. Google Cloud Spanner is the only cloud database providing external consistency (linearizability) across a globally distributed deployment as a standard feature. GCP is the only provider offering proprietary TPU (Tensor Processing Unit) hardware: TPU v5e, v5p, and the Ironwood (v7) inference-first chip provide AI training and inference acceleration unavailable from any other cloud. Gemini 2.5 Pro's 1-million-token context window and multimodal capabilities (text, image, audio, video via Imagen, Chirp, Veo) represent the most integrated first-party foundation model platform. GCP invented Kubernetes (open-sourced as Kubernetes), and GKE Autopilot remains the most fully automated serverless Kubernetes offering, with per-pod billing and enforced security baselines. Apigee is a leader in the 2025 Gartner Magic Quadrant for API Management, with enterprise API lifecycle management capabilities (monetization, API Hub governance, hybrid deployment) that surpass the simpler API gateways offered by AWS and OCI.

Choosing the Right Cloud

These recommendations are workload-pattern based, not absolute. Most enterprise deployments run on more than one cloud, with workloads placed on the provider best suited to each use case.

Oracle Database workloads: OCI is the native home for Oracle Database, including Autonomous Database, Oracle RAC, Exadata, GoldenGate, APEX, and ORDS. Oracle Database@Azure and Oracle Database@Google Cloud are available for organizations that need Oracle workloads co-located with Azure or GCP services, but those are Oracle-managed deployments running on Oracle hardware inside Azure/GCP data centers.

AI/ML training and inference at scale: GCP is the strongest choice when proprietary TPU hardware is required for training (JAX-based frameworks targeting TPU pods) or when BigQuery ML / Vertex AI integration is the target. AWS is the strongest choice for model diversity (Bedrock) and for managed training cluster operations (SageMaker HyperPod). OCI is the strongest choice for pure GPU cluster scale with bare metal access.

Microsoft ecosystem workloads: Azure is the natural home for workloads that depend on Microsoft Entra ID, Active Directory, Microsoft 365, Dynamics 365, SQL Server, or .NET. The Microsoft stack integrates at a depth unavailable on AWS, OCI, or GCP.

Data-intensive analytics: GCP BigQuery eliminates warehouse operations entirely and delivers per-query pricing that benefits variable workloads. AWS provides the deepest open lakehouse (Iceberg) ecosystem when multi-engine access across EMR, Athena, Redshift, and Glue is required.

Price-sensitive workloads: OCI's consistently lower pricing on compute (flexible shapes), egress, networking (free NAT, DRG, private endpoints), and object storage makes it competitive for workloads where data volume and networking costs are significant factors. OCI Preemptible Instances at a flat 50% discount with no bidding simplify cost-efficient batch and training jobs.

Regulated and compliance-critical workloads: Azure has the broadest compliance certification portfolio, deepest government cloud capabilities (Azure Government, M365 Government), and the only native multi-cloud governance plane (Azure Arc) for organizations managing compliance across heterogeneous environments. AWS GovCloud and GCP Assured Workloads serve comparable regulated workloads on their respective platforms.

Multi-cloud strategy: All four providers support Terraform as a first-class IaC tool (OCI and GCP both ship managed Terraform services as their native IaC). Kubernetes (GKE, EKS, AKS, OKE) provides a portable application runtime layer. The primary lock-in vectors are managed databases (proprietary engines like Aurora, Spanner, Autonomous DB), AI/ML platforms, and deep PaaS services (Lambda, Azure Functions, Cloud Run) — evaluate these dependencies explicitly when planning multi-cloud portability.