AZ-104 vs AI-200 — take AZ-104 if your work involves managing Azure infrastructure, configuring virtual networks, and administering cloud environments. Take AI-200 if your work involves building cloud-native applications that integrate AI services, containerized workloads, and event-driven pipelines on Azure. AZ-204 (Azure Developer Associate) retires July 31, 2026 and AI-200 is its official replacement — so if you were planning to take AZ-204, this comparison is exactly what you need to read right now.
These are not competing certifications at the same level of difficulty. They are two separate associate-level Azure paths for two completely different job roles.
AZ-104 vs AI-200: Key Differences at a Glance
| Factor | AZ-104 | AI-200 |
| Official name | Microsoft Azure Administrator | Azure AI Cloud Developer Associate |
| Replaces | Nothing — active, no retirement planned | AZ-204 (Azure Developer Associate, retiring July 31, 2026) |
| Exam cost | $165 USD | $165 USD |
| Exam duration | 120 minutes | 120 minutes |
| Passing score | 700 out of 1000 | 700 out of 1000 |
| Expiration | 1 year — free annual renewal | 1 year — free annual renewal |
| Prerequisites | None formally required | None formally required |
| Primary audience | System administrators, cloud engineers, IT operations | Cloud developers, backend engineers, AI solution builders |
| Primary focus | Configure, manage, and monitor Azure infrastructure | Build, integrate, and monitor AI solutions on Azure cloud infrastructure |
| Coding required | No — CLI commands tested but no programming | Yes — Python and cloud-native development experience expected |
| Live lab simulations | Yes — configure real Azure resources in exam | Not confirmed — scenario and configuration questions |
| Core platform | Azure Portal, Azure CLI, PowerShell | Azure Container Apps, Azure Functions, vector databases, Azure OpenAI |
| AI content | Minimal — Azure AI services overview only | Core — building AI-powered applications is the primary focus |
| Average US salary | $88,000 to $161,000 | $110,000 to $150,000 |
| Best for | Cloud administrators, infrastructure engineers, IT professionals | Cloud developers, AI application engineers, DevOps professionals |
| Leads to | AZ-305 (Architect Expert), SC-500 (Security), AZ-400 (DevOps) | AZ-400 (DevOps Expert), AI-103, AI-300 |
| Beta available | No — active since 2020 | April 2026 — general availability July 2026 |
The Critical 2026 Context: AZ-204 Is Retiring
Before choosing between AZ-104 and AI-200, you need to understand the most important fact about the Azure developer certification path in 2026.
AZ-204 (Azure Developer Associate) retires July 31, 2026. AI-200 (Azure AI Cloud Developer Associate) is its official replacement. The shift from AZ-204 to AI-200 moves from broad Azure application development to AI-focused cloud-native infrastructure specifically for AI solutions — containers, serverless functions, vector databases, and event-driven pipelines.
This means if you were planning to take AZ-204 as your Azure developer certification, you have two choices right now. Finish AZ-204 before July 31, 2026 if you are already well into preparation. Go directly to AI-200 if you are starting from scratch today. For the complete breakdown of that specific transition, our AZ-204 vs AI-200 guide covers every detail.
What Is the Main Difference Between AZ-104 and AI-200?
The core difference between AZ-104 and AI-200 is the difference between managing what exists and building what is new.
AZ-104 is an operations and administration certification. The AZ-104 certified professional configures the Azure environment that everyone else uses. They provision virtual machines, manage network security groups, set up storage accounts, implement RBAC policies, configure backup and recovery, and monitor the health of running systems. They keep Azure working.
AI-200 is a development and integration certification. The AI-200 certified professional builds applications that run on Azure. They deploy containerized workloads, write serverless functions, integrate vector databases for AI search, connect AI services to enterprise data pipelines, and implement observability across distributed systems. They build the things that run in the environment AZ-104 manages.
Both roles are essential in every cloud organization. They sit side by side. They rarely overlap significantly in their daily work.
What Does AZ-104 Cover?
AZ-104 is the Microsoft Azure Administrator Associate certification. It validates your ability to implement, manage, and monitor Azure infrastructure across the five core operational domains that define day-to-day Azure administration.
AZ-104 Exam Domains
| Domain | Weight | What You Do |
| Manage Azure identities and governance | 15-20% | Configure Microsoft Entra ID, manage users and groups, implement RBAC, manage subscriptions and resource groups, Azure Policy, resource locks |
| Implement and manage storage | 15-20% | Configure storage accounts, manage blob storage and Azure Files, implement storage redundancy and lifecycle policies, Azure Storage Explorer |
| Deploy and manage Azure compute resources | 20-25% | Deploy and configure VMs, manage VM scale sets, configure Azure App Service, Azure Container Instances, basic AKS |
| Implement and manage virtual networking | 15-20% | Configure VNets and subnets, NSGs, VPN Gateways, Azure DNS, Load Balancer, Application Gateway, VNet peering |
| Monitor and maintain Azure resources | 10-15% | Azure Monitor, Log Analytics, Azure Alerts, Azure Backup, Azure Site Recovery, Azure Update Manager |
AZ-104’s live lab requirement: AZ-104 includes live lab simulations where you configure real Azure resources during the exam using the actual Azure Portal. You cannot pass through theory alone. Candidates must have genuine hands-on experience building and configuring Azure environments. For the complete AZ-104 salary and career analysis, our Is AZ-104 Worth It guide covers every detail including the $88,000 to $161,000 salary range and 78,000 plus active job openings.
What Does AI-200 Cover?
AI-200 is the Azure AI Cloud Developer Associate certification. It validates your ability to build, integrate, and monitor AI solutions on Azure using cloud-native infrastructure — containers, serverless, vector databases, event-driven pipelines, and distributed observability.
AI-200 Exam Domains
Based on Microsoft’s official AI-200 study guide and course description:
| Domain | What You Build |
| Plan and manage AI resources | Plan Azure AI solution architecture, select compute and data services for AI workloads, manage AI resource lifecycle |
| Build AI solutions on Azure | Deploy containerized applications to Azure Container Apps and AKS, build serverless APIs with Azure Functions, implement event-driven architectures with Service Bus and Event Grid |
| Work with vector-enabled data | Design and query solutions with Cosmos DB for NoSQL, PostgreSQL with pgvector, Azure Managed Redis for vector search and semantic similarity |
| Implement AI pipelines and integrations | Build event-driven AI pipelines, connect AI services to enterprise data, implement RAG patterns with Azure-native data services |
| Secure and observe AI solutions | Implement secrets management with Key Vault, configure managed identities, implement distributed observability with Azure Monitor and Application Insights |
Core tools tested: Azure Container Apps, Azure Kubernetes Service, Azure Functions, Azure Service Bus, Azure Event Grid, Cosmos DB for NoSQL, Azure Database for PostgreSQL with pgvector, Azure Managed Redis, Azure Key Vault, Azure Monitor, Application Insights. For the full comparison of AI-200 and AI-103 covering the two main new Azure AI developer certifications, our AI-103 vs AI-200 guide covers every difference.
AZ-104 vs AI-200: Difficulty Comparison
| Factor | AZ-104 | AI-200 |
| Technical background needed | IT administration, systems engineering | Software development, cloud-native application building |
| Hardest area | Virtual networking and live lab simulations | Vector database integration and distributed observability |
| Study time | 6 to 10 weeks with Azure experience | 6 to 10 weeks with development and Azure experience |
| Coding required | No — CLI commands recognized but no programming | Yes — Python and application development experience expected |
| Lab simulations | Yes — configure real Azure Portal during exam | Not confirmed in beta |
| Common failure reason | Insufficient hands-on networking practice | Insufficient experience with vector databases and containerization |
| Recommended experience | 6 months Azure administration | 1 year cloud-native application development |
AZ-104 is harder in terms of the live lab simulation requirement — you must be able to configure real Azure resources under timed exam pressure. AI-200 is harder in terms of breadth of technical skills required — you need genuine development experience, containerization knowledge, and understanding of AI integration patterns before the content becomes intuitive.
Difficulty is determined primarily by your background. An experienced Azure administrator finds AZ-104 content natural and AI-200 content foreign. An experienced cloud developer finds AI-200 content natural and AZ-104’s infrastructure focus less familiar.
AZ-104 vs AI-200: Salary Comparison
AZ-104 Salary by Role
| Role | Average US Salary |
| Azure Administrator | $95,000 to $130,000 |
| Cloud Engineer | $100,000 to $135,000 |
| Systems Engineer (Azure) | $93,000 to $167,000 |
| Infrastructure Engineer | $95,000 to $125,000 |
| Senior Cloud Engineer | $145,000 to $218,000 |
AI-200 Salary by Role
| Role | Average US Salary |
| Azure AI Cloud Developer | $110,000 to $145,000 |
| Cloud Software Engineer (AI focus) | $115,000 to $150,000 |
| Backend Developer (cloud-native) | $105,000 to $140,000 |
| AI Platform Engineer | $120,000 to $155,000 |
| Senior Cloud AI Developer | $140,000 to $180,000 |
AI-200 roles typically command a salary premium over AZ-104 roles at comparable experience levels. This reflects the greater technical complexity of cloud-native AI development compared to infrastructure administration and the scarcity of professionals who combine application development skills with AI integration knowledge.
However AZ-104 opens significantly more total job postings — 78,000 plus active positions versus AI-200’s narrower but growing developer role market. AZ-104 is the higher-volume opportunity. AI-200 is the higher-salary opportunity for professionals with the matching development background.
Who Should Take AZ-104?
Take AZ-104 if:
You are an IT administrator, systems engineer, or infrastructure professional. AZ-104 validates the skills you already use if you manage compute, storage, networking, and identity in Azure environments. It is the most directly applicable certification for operations-focused cloud professionals.
Your career goal is cloud architecture through the AZ-305 path. AZ-104 is a mandatory prerequisite for AZ-305 (Azure Solutions Architect Expert). You cannot earn the architect certification without an active AZ-104. Every professional targeting cloud architecture must hold AZ-104 first. For the complete analysis of that progression, our AZ-104 vs AZ-305 guide covers when to make the jump.
You want maximum Azure follow-on certification flexibility. AZ-104 is required or recommended before AZ-305, AZ-400, and multiple specialty certifications. It is the broadest prerequisite-generating Azure associate certification available.
Your background is IT operations rather than software development. AZ-104 does not require coding. It tests operational skills — configuring, troubleshooting, monitoring — rather than programming. Infrastructure professionals without development backgrounds find AZ-104 significantly more approachable than AI-200.
Who Should Take AI-200?
Take AI-200 if:
You are a software developer or backend engineer building applications on Azure. AI-200 validates cloud-native development skills — containerization, serverless APIs, event-driven pipelines, vector databases — that developers building AI-powered applications on Azure use daily.
You were planning to take AZ-204 and can no longer start preparation before July 31. AZ-204 retires July 31, 2026. Candidates who are starting fresh on the Azure developer path should go directly to AI-200 rather than investing preparation time in a certification with less than 3 months of remaining availability.
Your work involves integrating AI services into cloud applications. AI-200 specifically validates the ability to connect AI services to enterprise data, implement vector search, build RAG pipelines on Azure-native data services, and observe AI workloads. If this describes your daily work, AI-200 is the most direct credential for your role.
You want to build toward AZ-400 DevOps Engineer Expert through a developer path. AZ-400 accepts either AZ-104 or AI-200 as a qualifying prerequisite. Developers who earn AI-200 and then pursue AZ-400 build a developer-plus-DevOps credential profile that is particularly strong for cloud engineering leadership roles.
Can You Take Both AZ-104 and AI-200?
Yes — and many senior cloud engineers eventually hold both. The combination validates complete Azure platform competency from infrastructure administration through cloud-native AI application development.
Organizations deploying AI-powered applications on Azure need professionals who can both build the application and manage the infrastructure it runs on. Professionals who hold both AZ-104 and AI-200 can serve this full-stack operational and development need. This profile is particularly valuable in smaller organizations where infrastructure and development responsibilities overlap, and in consulting practices where client engagements span both layers.
The most common sequence is AZ-104 first for infrastructure professionals who then add AI-200 to develop cloud-native application skills. For developers, the reverse sequence — AI-200 first then AZ-104 — builds infrastructure understanding on top of existing application development expertise.
The Complete Azure Associate Certification Landscape in 2026
Understanding where AZ-104 and AI-200 fit among all current Azure associate certifications helps you plan your full certification roadmap.
| Certification | Role | Focus | Primary Background |
| AZ-104 | Azure Administrator | Configure and manage Azure infrastructure | IT administration, systems engineering |
| AI-200 (new July 2026) | Azure AI Cloud Developer | Build AI solutions on Azure cloud infrastructure | Software development, cloud engineering |
| AI-103 (new June 2026) | Azure AI Apps and Agents Developer | Build generative AI apps and agents via Foundry | AI application development |
| AI-300 (new May 2026) | MLOps Engineer | Operationalize ML and GenAI in production | Data science, MLOps engineering |
| DP-700 | Fabric Data Engineer | Build data pipelines on Microsoft Fabric | Data engineering |
| DP-750 | Azure Databricks Data Engineer | Build data pipelines on Databricks | Big data engineering |
| SC-500 (new July 2026) | Cloud and AI Security Engineer | Secure cloud and AI environments | Security engineering |
For the complete picture of every Microsoft certification change happening in 2026 including every retirement date and replacement, our Microsoft certifications retiring in 2026 guide covers every deadline and action plan. For the complete Azure certification path from AZ-900 through to expert level, our AZ-104 vs AZ-900 guide covers the foundational decision every Azure career starts with.
How to Prepare for AZ-104
Step 1: Build hands-on Azure experience from day one. AZ-104 includes live lab simulations — you configure real Azure resources in the actual Azure Portal under exam conditions. Create a free Azure account and start building immediately. Deploy virtual machines, create virtual networks, configure storage accounts, and implement RBAC policies. Hands-on experience is mandatory for AZ-104 success.
Step 2: Master virtual networking specifically. Virtual networks, subnets, NSGs, VPN Gateways, Load Balancers, and DNS configuration are consistently the most heavily tested and most commonly failed areas of AZ-104. Build multiple VNet configurations hands-on. Understand traffic flow and security rule evaluation deeply.
Step 3: Learn PowerShell and Azure CLI for key tasks. AZ-104 tests CLI knowledge for common administrative tasks. Run every CLI command you study in a real Azure terminal rather than just reading about it.
Step 4: Use current practice materials. Our Microsoft exam preparation section covers current AZ-104 practice materials.
How to Prepare for AI-200
Step 1: Build genuine cloud-native development experience first. AI-200 expects candidates who write code and deploy applications. If containerization, serverless functions, and event-driven architectures are not part of your daily work, build that experience before attempting AI-200 preparation.
Step 2: Get hands-on with vector databases on Azure. Vector database integration is a distinctive new content area for AI-200. Practice with Cosmos DB for NoSQL vector search, Azure Database for PostgreSQL with the pgvector extension, and Azure Managed Redis. Build a semantic search scenario on each platform.
Step 3: Build an event-driven AI pipeline. The event-driven architecture domain is central to AI-200. Build a pipeline where an Azure Function triggers on an event, calls an Azure AI service, stores results in a vector database, and serves downstream AI applications.
Step 4: Study the official AI-200T00 course content. Microsoft’s official course AI-200T00: Develop AI Cloud Solutions on Microsoft Azure launches alongside general availability in July 2026 and covers the full exam scope including containerization patterns, serverless APIs, and vector-enabled data services.
Step 5: Use current practice materials. Our Microsoft exam preparation section covers current active Microsoft certifications.
Decision Framework: AZ-104 vs AI-200
| Your Situation | Take This |
| You are a systems administrator or IT operations professional | AZ-104 |
| You are a software developer or backend engineer | AI-200 |
| You configure and manage Azure infrastructure daily | AZ-104 |
| You build and deploy cloud-native applications | AI-200 |
| You want to become a cloud architect | AZ-104 — required for AZ-305 |
| You want to become a cloud-native AI developer | AI-200 |
| You need AZ-305 as a prerequisite | AZ-104 — AZ-305 requires it |
| You need AZ-400 as a prerequisite | Either — both qualify |
| You come from a no-code IT background | AZ-104 |
| You come from a software development background | AI-200 |
| You were planning AZ-204 but cannot finish before July 31 | AI-200 directly |
| You already hold AZ-104 and want developer skills | AI-200 |
| You already hold AI-200 and want infrastructure skills | AZ-104 |
| Maximum job posting access | AZ-104 — 78,000 plus active postings |
| Maximum salary ceiling | AI-200 — senior AI developer roles pay more |
Frequently Asked Questions: AZ-104 vs AI-200
What is the difference between AZ-104 and AI-200?
AZ-104 validates Azure infrastructure administration skills — configuring and managing virtual machines, networking, storage, identity, and monitoring. AI-200 validates cloud-native AI development skills — building containerized applications, serverless APIs, vector databases, and event-driven AI pipelines on Azure. AZ-104 is for administrators. AI-200 is for developers.
Is AI-200 replacing AZ-204?
Yes. AZ-204 (Azure Developer Associate) retires July 31, 2026 and AI-200 (Azure AI Cloud Developer Associate) is its official replacement. The shift moves from broad Azure application development to AI-focused cloud-native infrastructure specifically for AI solutions.
Which is harder — AZ-104 or AI-200?
Both are intermediate level but challenging in different ways. AZ-104 requires genuine hands-on Azure infrastructure experience and includes live lab simulations. AI-200 requires genuine cloud-native development experience with containers, serverless, and vector databases. Difficulty depends entirely on your background.
Which pays more — AZ-104 or AI-200?
AI-200 pays more at comparable experience levels. AI-200 roles average $110,000 to $150,000. AZ-104 roles average $88,000 to $130,000. However AZ-104 opens significantly more total job postings — 78,000 plus — while AI-200 targets a narrower, higher-paying developer market.
Does AZ-104 require coding?
No. AZ-104 does not require programming. It tests CLI knowledge for administrative tasks but no application development or programming skills.
Does AI-200 require coding?
Yes. AI-200 targets software developers and expects genuine programming experience, particularly in Python and cloud-native development patterns including containerization and serverless functions.
Can AI-200 satisfy the AZ-305 prerequisite?
No. AZ-305 (Azure Solutions Architect Expert) requires an active AZ-104 as its prerequisite. AI-200 does not satisfy the AZ-305 prerequisite requirement. Candidates targeting cloud architecture must earn AZ-104 before pursuing AZ-305.
Can AI-200 satisfy the AZ-400 prerequisite?
Yes. AZ-400 (Azure DevOps Engineer Expert) accepts either AZ-104 or AI-200 as a qualifying prerequisite. This makes AI-200 a valid path to the DevOps Expert certification for developers who prefer not to take AZ-104 first.
When is AI-200 available?
AI-200 went into beta in April 2026 and reaches general availability in July 2026. The official course AI-200T00 launches alongside general availability in July 2026.
What is the difference between AI-200 and AI-103?
AI-200 validates cloud infrastructure skills for building AI solutions — containers, serverless, vector databases, event-driven pipelines. AI-103 validates AI application and agent development skills — building generative AI apps and autonomous agents using Microsoft Foundry. AI-200 focuses on the cloud infrastructure layer. AI-103 focuses on the AI application layer. For the complete comparison, our AI-103 vs AI-200 guide covers every difference.
