AI & Machine Learning: AWS to Azure service map
65 AWS → Azure mappings: 34 direct, 21 partial, 10 with no equivalent.
Foundation models, agents, RAG, ML platforms, MLOps, AI chips, assistants, and prebuilt AI services. AWS mapped to Azure, verified against official docs.
Generative AI platforms & models(8)
Foundation model API platform Direct match
The Azure equivalent of Amazon Bedrock is Microsoft Foundry Models (which includes Azure OpenAI). Both are managed, pay-per-token, multi-vendor model APIs.
Cloud provider's own foundation models Partial match
Amazon Nova's closest Azure counterpart is Microsoft's own models in Foundry (MAI and Phi, plus Model Router). But Microsoft's flagship LLMs on Azure are mainly OpenAI GPT models, so there's no one-to-one family match.
Open-weight model hub on dedicated compute Direct match
SageMaker JumpStart and Bedrock Marketplace map to the Microsoft Foundry model catalog's managed-compute deployments (and the Azure Machine Learning model catalog).
Foundation model fine-tuning & distillation Direct match
Bedrock model customization maps to Foundry fine-tuning. Both offer supervised fine-tuning, reinforcement fine-tuning, and distillation for selected models.
Reserved model throughput & cross-region routing Direct match
Bedrock Provisioned Throughput and cross-Region inference map to Foundry Provisioned (PTU) deployments and the Global/Data Zone deployment types.
Prompt & LLM workflow orchestration Partial match
Bedrock Flows (with Prompt management) maps most closely to Microsoft Agent Framework workflows on Azure. The older Azure Prompt flow is deprecated and retires on April 20, 2027.
Model, RAG & agent evaluation Direct match
Bedrock Evaluations and AgentCore Evaluations map to Foundry observability and evaluations (built-in and custom evaluators, continuous evaluation, and tracing).
GenAI multimodal content extraction (IDP) Direct match
The Azure equivalent of Amazon Bedrock Data Automation is Azure Content Understanding.
AI agents(6)
Managed AI agent platform Direct match
The Azure equivalent of Amazon Bedrock AgentCore is Foundry Agent Service. Both host declarative or code-based agents from any framework and model.
Configuration-based agents (legacy Bedrock Agents) Partial match
Bedrock Agents Classic is in maintenance (no new customers since July 30, 2026). On Azure, the equivalent concept is a Foundry prompt agent, and on AWS the successor is the AgentCore managed harness.
Agent tools, MCP gateway & sandboxes Partial match
AgentCore Gateway, Code Interpreter, and Browser map to the Foundry Agent Service toolbox (a managed MCP endpoint with web search, file search, code interpreter, OpenAPI, functions, and MCP tools).
Agent memory (short- & long-term) Partial match
AgentCore Memory maps to Foundry Agent Service memory (preview). AgentCore's memory is GA and more configurable today.
Agent identity, policy & registry Partial match
AgentCore Identity, Policy, and Registry map to Microsoft Entra agent identities, Foundry guardrails (tool-call controls), the Foundry control plane, and the Entra Agent Registry.
Open-source agent SDK from the cloud vendor Direct match
The Azure counterpart of the Strands Agents SDK is Microsoft Agent Framework. Both are open-source, code-first agent SDKs that deploy to the vendor's managed agent runtime (AgentCore Runtime or Foundry hosted agents).
Safety, RAG & search(3)
GenAI guardrails & content safety Direct match
The Azure equivalent of Amazon Bedrock Guardrails is Foundry guardrails, which are powered by Azure AI Content Safety. Content Safety is also available as a standalone API, like Bedrock's ApplyGuardrail.
Managed RAG / knowledge bases Direct match
The Azure equivalent of Amazon Bedrock Knowledge Bases is Foundry IQ knowledge bases, which run on Azure AI Search with agentic retrieval.
Enterprise / intelligent search Partial match
Amazon Kendra maps to Azure AI Search, but Kendra is in maintenance mode (no new customers since July 30, 2026), and AWS now points new search workloads to Bedrock Managed Knowledge Base.
ML platform (SageMaker AI ↔ Azure ML)(9)
End-to-end ML platform Direct match
The Azure equivalent of Amazon SageMaker AI is Azure Machine Learning.
Unified data + analytics + AI workspace Partial match
Amazon SageMaker Unified Studio maps most closely to Microsoft Fabric (Data Science workload) integrated with Azure Machine Learning. Microsoft's own comparison maps SageMaker to the Fabric Data Science workload plus Azure ML.
ML IDE & managed notebooks Direct match
Amazon SageMaker Studio maps to Azure Machine Learning studio with compute instances (JupyterLab and VS Code).
Managed training jobs Direct match
SageMaker AI training jobs map to Azure Machine Learning command/sweep jobs on compute clusters or serverless compute.
Resilient large-scale FM training clusters Partial match
Azure has no single managed equivalent of SageMaker HyperPod. The usual Azure patterns are Azure ML compute clusters/serverless with ND-series GPUs, or AKS with GPU node pools attached to Azure ML.
Model hosting & inference endpoints Direct match
SageMaker AI endpoints map to Azure Machine Learning managed online endpoints (real-time), batch endpoints, and serverless deployments.
AutoML Direct match
SageMaker Autopilot (now surfaced in SageMaker Canvas) maps to Azure Machine Learning automated ML.
No-code ML for analysts Partial match
SageMaker Canvas maps partially to the no-code AutoML UI and Designer in Azure Machine Learning studio. Azure has no single analyst-focused product with the same scope.
Preconfigured deep learning VM images Direct match
The Azure equivalent of AWS Deep Learning AMIs is Data Science Virtual Machines.
MLOps & responsible AI(9)
ML pipelines / workflow orchestration Direct match
Amazon SageMaker Pipelines maps to Azure Machine Learning pipelines.
Feature store Direct match
Amazon SageMaker Feature Store maps to the Azure Machine Learning managed feature store.
Model registry Direct match
SageMaker Model Registry maps to the Azure Machine Learning model registry and cross-workspace registries.
Experiment tracking (MLflow) Direct match
Managed MLflow on SageMaker AI maps to MLflow in Azure Machine Learning. On Azure, every workspace is an MLflow-compatible tracking server.
Bias detection & explainability Partial match
SageMaker Clarify (in maintenance, no new customers) maps to the Azure Machine Learning Responsible AI dashboard. On AWS, new customers are directed to SHAP, MLflow, and Bedrock Evaluations instead.
Production model monitoring (drift) Direct match
SageMaker Model Monitor (in maintenance) maps to Azure Machine Learning model monitoring. Azure's is a fully supported, built-in service, while AWS now recommends open-source monitoring solutions for new customers.
Data labeling (human + ML-assisted) No equivalent
Azure no longer has a first-party equivalent of SageMaker Ground Truth. Azure Machine Learning data labeling retired on September 30, 2026, and Microsoft points to third-party labeling providers. Ground Truth itself is in maintenance (no new customers).
Human-in-the-loop review of predictions No equivalent
Azure has no direct equivalent of Amazon Augmented AI (A2I), and A2I itself is in maintenance (no new customers since the June 30, 2026 announcement).
Training debugging & profiling No equivalent
Azure Machine Learning has no managed equivalent of SageMaker Debugger and Profiler. On AWS, Debugger is in maintenance and Profiler reaches end of support on June 30, 2027.
AI infrastructure & chips(3)
NVIDIA GPU instances for AI training Direct match
Amazon EC2 P-family instances (P5/P5e/P5en, P6-B200, P6-B300, P6e-GB200) map to Azure ND-series VMs (ND H100 v5, ND H200 v5, ND GB200 v6, ND GB300 v6).
Custom AI training silicon (rentable) No equivalent
Azure has no rentable equivalent of AWS Trainium. Microsoft's Maia 200 is an inference accelerator used behind Microsoft Foundry and Microsoft 365 Copilot, not sold as VMs (the SDK is in preview).
Custom AI inference silicon (rentable) No equivalent
Azure has no rentable equivalent of AWS Inferentia (Inf2). Microsoft's Maia 200 inference chip isn't offered as customer VMs.
AI assistants & copilots(3)
Enterprise AI work assistant Partial match
Microsoft 365 Copilot is the Microsoft counterpart of Amazon Q Business. Q Business is in maintenance (no new customers), and AWS's successor is Amazon Quick.
AI coding assistant / agentic IDE Direct match
GitHub Copilot is Microsoft's equivalent of Amazon Q Developer in the IDE. AWS is replacing the Q Developer IDE plugins with Kiro, and the plugins reach end of support on April 30, 2027.
Cloud console / operations assistant Direct match
The Azure equivalent of Amazon Q Developer in the AWS Management Console is Azure Copilot.
Vision & documents(7)
Prebuilt image analysis Partial match
Amazon Rekognition's image features map partly to Azure Vision Image Analysis, but that API is deprecated and retires on September 25, 2028. Microsoft now recommends Foundry multimodal models and Content Understanding instead.
Face detection, verification & search Partial match
Rekognition face APIs map to Azure Face. Azure Face recognition is Limited Access, so you have to apply and be a Microsoft-managed customer or partner.
Video analysis & indexing Partial match
Amazon Rekognition Video maps to Azure AI Video Indexer. Video Indexer is a broader "insights" application (transcripts, translation, topics, generative summaries) rather than a low-level detection API.
Image & text content moderation Direct match
Rekognition content moderation (and Comprehend toxicity detection for text) maps to Azure AI Content Safety.
Custom image classification & detection Partial match
Rekognition Custom Labels maps to Azure Custom Vision, but Custom Vision is retiring (supported until September 25, 2028). Microsoft recommends Azure ML AutoML for images, Foundry models, or Content Understanding for new work.
Visual defect / anomaly detection (manufacturing) No equivalent
Neither cloud has a current dedicated service. Amazon Lookout for Vision shut down on October 31, 2025, and Azure Custom Vision, the closest Azure tool, is itself retiring.
Document OCR & form/table extraction Direct match
The Azure equivalent of Amazon Textract is Azure Document Intelligence.
Language & translation(5)
Text analytics (entities, sentiment, key phrases) Direct match
The Azure equivalent of Amazon Comprehend is Azure Language. Note that Azure now labels sentiment, key phrases, summarization, and entity linking as "legacy" capabilities, supported for existing implementations.
Custom text classification & entity recognition Direct match
Comprehend custom classification and custom entity recognition map to Azure Language custom text classification (legacy) and custom NER (core).
PII detection & redaction in text Direct match
Comprehend PII detection maps to Azure Language PII detection, which is a "core" Azure Language capability covering text, conversations, and documents.
Machine translation Direct match
The Azure equivalent of Amazon Translate is Azure Translator.
Medical text analytics Direct match
The Azure equivalent of Amazon Comprehend Medical is Text Analytics for health in Azure Language.
Speech & conversational AI(5)
Speech to text Direct match
The Azure equivalent of Amazon Transcribe is Azure Speech speech-to-text.
Text to speech Direct match
The Azure equivalent of Amazon Polly is Azure Speech text-to-speech.
Contact-center call analytics Partial match
Amazon Transcribe Call Analytics maps partially to Azure Speech batch transcription combined with Azure Language or Foundry models. Microsoft documents this as a call-center scenario rather than a single API.
Real-time speech-to-speech AI Partial match
Amazon Nova Sonic (speech-to-speech on Bedrock) maps to Azure Voice Live and voice-based prompt agents in Foundry (preview), which also support realtime audio models.
Conversational bot builder (intents & dialog) Partial match
Amazon Lex maps to Microsoft Copilot Studio (low-code agents) together with Azure AI Bot Service (channels). Azure's intent engine, CLU, is deprecated and retires March 31, 2029.
Industry & specialized ML(7)
Managed recommendation engine No equivalent
Azure has no current equivalent of Amazon Personalize. Azure AI Personalizer retired on August 25, 2026. On Azure, you build recommenders yourself in Azure Machine Learning or Fabric.
Time-series forecasting Partial match
Amazon Forecast (no new customers since July 2024) maps to Azure Machine Learning AutoML forecasting. AWS itself points Forecast users to SageMaker Canvas.
Managed fraud detection No equivalent
Azure has no managed fraud-detection API comparable to Amazon Fraud Detector, and Fraud Detector itself closed to new customers on November 7, 2025. Both clouds now point to building models (AutoGluon on SageMaker AI, or Azure ML AutoML).
Industrial equipment anomaly detection / predictive maintenance No equivalent
There's no current managed service on either side. Amazon Lookout for Equipment reached end of support on October 7, 2026, Amazon Monitron is closed to new customers, and Azure AI Anomaly Detector retired on October 1, 2026.
Business-metric anomaly detection No equivalent
There's no current equivalent on either side. Amazon Lookout for Metrics shut down on October 10, 2025, and Azure AI Metrics Advisor retired on May 18, 2026.
FHIR health data store Direct match
The Azure equivalent of AWS HealthLake is the FHIR service in Azure Health Data Services. The older Azure API for FHIR was deprecated on September 30, 2026.
Ambient clinical documentation (AI scribe) Partial match
Azure has no documented one-API equivalent of AWS HealthScribe. You can compose it from Azure Speech, Text Analytics for health, and Foundry models (including Microsoft healthcare AI models).