AI-200 Study Guide: Microsoft Azure AI Cloud Developer Associate (2026)

AI-200 is the exam for Microsoft Certified: Azure AI Cloud Developer Associate, Microsoft’s developer certification for building the back end of AI solutions on Azure: containers, vector-enabled databases, event-driven services, serverless functions, secrets and monitoring. It costs $165 in the US for a 120-minute exam with a passing score of 700, and it replaced AZ-204 (Azure Developer Associate), which Microsoft has retired.

If you’re coming from AZ-204, much of the ground is familiar, but the storage section has become a 25–30% block on building AI features with Azure Cosmos DB, Azure Database for PostgreSQL and Azure Managed Redis, including vector search and retrieval-augmented generation (RAG). This guide covers the full cost, the four skill areas and weights, how AI-200 compares with AZ-204 and AI-103, a 6-week study plan and the free official prep. Microsoft has no AI-200 Practice Assessment yet.

AI-200 at a glance

Checked against Microsoft Learn’s Azure AI Cloud Developer Associate certification page, AI-200 study guide and Azure Developer Associate page on October 8, 2026.

How much the AI-200 certification costs

The exam fee is $165 in the US. Microsoft’s exam pricing data lists the same tier at £106 in the UK, €126 in Germany and France, ₹4,865 in India, ¥20,300 in Japan and US$140 in Australia and Canada, before tax. Microsoft notes that listed prices don’t include promotional offers or the reduced pricing for Microsoft Certified Trainers and Microsoft partner program members.

So the realistic total is $165 if you pass the first time and $330 if you need a second attempt, plus tax. Our AI certification cost calculator and AI certification cost guide compare this with other vendors.

AI-200 replaced AZ-204: what changed

Microsoft’s AZ-204 page now says the certification and its renewal assessment are retired, and Microsoft’s retirement announcement lists July 31, 2026 as the retirement date, with AI-200 as the replacement. The same announcement says certifications you already earned stay valid until they expire. Here’s a rough mapping by area name:

Microsoft’s announcement describes AI-200 as validating your ability to build, integrate and monitor AI solutions with containerized compute, vector-enabled databases, event-driven AI pipelines, serverless functions, secret management and distributed observability. If you hold AZ-204, your certification stays valid until its expiry date, but its renewal assessment is retired, so AI-200 is Microsoft’s recommended route to a current certification for the same role.

AI-200 skills measured and weights

The study guide shows no “as of” date and was last updated May 5, 2026. Microsoft notes that the bullets under each skill illustrate how it is assessed, that related topics may be covered, and that most questions cover generally available features. Note what isn’t here: building generative AI apps and agents is AI-103’s job, and running machine learning models in production is AI-300’s.

Skill area 1: Develop containerized solutions on Azure (20–25%)

• Container hosting: building, storing, versioning and managing images in Azure Container Registry, building and running images with ACR Tasks, and deploying containers to App Service with environment variables and secrets.
• Container orchestration: deploying to Azure Container Apps (environments and revisions), event-driven scaling with KEDA in Container Apps, deploying to AKS with manifest files, and troubleshooting AKS and Container Apps through logs, events and end-to-end connectivity.

How to study it: containerize one small Python API and take it through the whole path: push it to Container Registry, run it on App Service, then on Container Apps with a scale rule, then on AKS with a manifest. Break something on purpose each time and find it in the logs.

Skill area 2: Develop AI solutions by using Azure data management services (25–30%)

• Azure Cosmos DB for NoSQL: connecting and querying with the SDK, tuning queries and Request Unit consumption with indexing policies and consistency levels, storing embeddings and running vector similarity search, and handling new or updated items with the change feed processor.
• Azure Database for PostgreSQL: connecting and querying with SDKs, schema and data type design, indexing strategies that cut query latency and pgvector compute overhead, sizing compute, memory and storage for vector workloads, vector similarity search and RAG patterns with metadata filters, and connection optimization.
• Azure Managed Redis: caching, expiration and invalidation, and vector indexing for similarity search.

How to study it: this is the heaviest area and the most “AI” part of the exam. Take one small set of documents, create embeddings, and store and search them three ways: in Cosmos DB, in PostgreSQL with pgvector, and in Redis. Compare how each one indexes vectors and filters by metadata, then add a Redis cache in front of the slowest query. If RAG is new to you, our guide to improving AI accuracy explains why grounding answers in your own data matters.

Skill area 3: Connect to and consume Azure services (20–25%)

• Event- and message-based AI solutions: queuing and processing back-end work with Azure Service Bus (messages, topics, subscriptions and dead-letter queues), and event-driven workflows with Azure Event Grid (filters, custom events and retries).
• Azure Functions: serverless APIs with triggers and bindings, and configuring and deploying function apps.

How to study it: wire your API from area 1 to a Service Bus queue and an Event Grid topic, then add a function that processes messages and sends failures to the dead-letter queue. Our piece on giving AI access to your systems through APIs covers why this plumbing matters for AI apps.

Skill area 4: Secure, monitor, and troubleshoot Azure solutions (20–25%)

• Security: storing, rotating and retrieving secrets with Azure Key Vault, and storing and retrieving configuration with Azure App Configuration.
• Monitoring and troubleshooting: tracing distributed systems with OpenTelemetry SDKs, and writing KQL queries to analyze logs and metrics.

How to study it: move every secret in your practice app into Key Vault and every setting into App Configuration, add OpenTelemetry tracing across the API, queue and function, then write five KQL queries that answer real questions: slowest requests, failed messages, error rate by hour.

AI-200 vs AI-103: which Azure AI developer exam?

Take AI-200 if you build and run the services an AI app depends on; take AI-103 if you build the generative AI and agent features themselves. Some developers will want both, and AI-103 is also one of the associate certifications that count toward Microsoft’s AB-100 expert certification. Our Microsoft AI certifications guide shows the full path, and AI-901 vs AI-103 helps if you’re earlier in the journey.

A 6-week AI-200 study plan

This plan assumes you write Python and have deployed something to Azure before, and can give AI-200 about 6 to 8 hours a week. Microsoft’s nine learning paths alone add up to more than 30 hours, so build in lab time from week 1.

Prefer a plan built around your own calendar? The free AI certification study plan generator makes one.

Free official AI-200 prep

• AI-200 study guide: the full outline plus links to documentation for Container Registry, App Service, Functions, Cosmos DB, Key Vault, Event Grid, Service Bus and more.
• Course AI-200T00-A: Develop AI cloud solutions on Azure: 5 days, instructor-led. Microsoft describes it as covering Azure compute and containers, serverless APIs with Azure Functions, Service Bus and Event Grid, Cosmos DB for NoSQL, PostgreSQL with pgvector, and Azure Managed Redis. The self-paced version is the nine learning paths in the plan above.
• Exam sandbox: a demo of the Microsoft exam interface and question types.
• Exam Readiness Zone: Microsoft’s exam-prep video series, linked from the study guide.
• No Practice Assessment yet. The certification page says it isn’t available and that Practice Assessments usually arrive within 8 weeks of an exam leaving beta and becoming generally available. Check the page again before you book.

AI-200 practice questions

HOW TO // AI doesn’t have AI-200 practice questions yet, and Microsoft’s free Practice Assessment isn’t out yet either, so your lab is your practice test: if you can build and debug every bullet in the outline, you’re ready. When the Practice Assessment appears on the certification page, take it early and again before you book.

If you’re also taking AI-103, we do have free AI-103 practice questions written from Microsoft’s published outline. For every exam we cover, see AI Exam Prep, and use the AI certification index for verified facts on other AI exams. Developers using GitHub Copilot may also want our GitHub Copilot (GH-300) study guide.

Skip any site selling “real” AI-200 questions. Those are dumps, and for an exam this new they are mostly guesswork.

What exam day looks like

• 120 minutes, proctored by Pearson VUE, online or at a test center. The exam may include interactive components.
• Microsoft Learn is open during the exam. On associate and expert exams you can open Learn in a split screen. No extra time is added, the clock keeps running, and Q&A, Practice Assessments and your profile are blocked.
• Breaks: five minutes of break time are built in, the clock keeps running while you are away, and you can’t return to questions you saw before the break.
• Register with a personal Microsoft account. Microsoft warns that exam records tied to a work or school account are lost if you leave that organization.

Source for the timing, break and Learn-access rules: Microsoft’s exam duration and exam experience page; retake rules from the retake policy.

Renewal: AI-200 is valid for one year

Microsoft’s credential expiration policy says associate certifications are valid for one year. To keep AI-200, pass the free, online, open-book renewal assessment on Microsoft Learn; the window opens six months before your expiry date and attempts are unlimited. Microsoft offers renewal once a certification has been available for more than six months. Our guide to which AI certifications expire compares renewal rules across vendors.

Is AI-200 worth it?

If you build back-end services on Azure, yes: it is the certification Microsoft now recommends in place of AZ-204, and its biggest area covers vector search and RAG on Cosmos DB, PostgreSQL and Redis. It is not an entry point to AI, though. Microsoft expects working Python and Azure SDK skills, and the exam is about services and plumbing rather than models. If you’re newer to AI, start with AI-901 (Azure AI Fundamentals) or see AI certifications for beginners; for the full ranking, see the best AI certifications.

HOW TO // AI is not affiliated with or endorsed by Microsoft. AI-200, AZ-204, Microsoft Azure, Azure Cosmos DB, Azure Kubernetes Service and Microsoft Foundry are trademarks of Microsoft Corporation; we reference them descriptively. All content is original and based on Microsoft’s published study guide; we do not reproduce or reveal real exam questions. Check the official certification page before you book.

Related guides

• AI Certification Cost Calculator (Free)
• How Much Do AI Certifications Cost? Every 2026 Exam, Retake and Renewal Price
• How to Improve AI Accuracy: 7 Fixes That Actually Work
• APIs Before Agents: Giving AI Access to Your Systems Safely
• AI-103 Study Guide & Cheat Sheet (Azure AI Apps & Agents Developer)