Microsoft Certified: Azure AI Engineer Associate (AI‑102) Practice Exams
About the Azure AI-102 exam
AI-102 retired on 30 June 2026 — AI-103 is the current exam
Microsoft retired AI-102 and the Azure AI Engineer Associate certification on 30 June 2026, and lists AI-103 Azure AI Apps and Agents Developer Associate as its replacement. Anyone already certified keeps the credential on their transcript until it expires; it can no longer be renewed, and the exam can no longer be booked. New candidates take AI-103 instead. AI-103 is a separate exam rather than a reissue of AI-102 — it is built around Microsoft Foundry, generative apps and production agents — so check its own skills-measured list rather than assuming AI-102 preparation carries over. These practice questions target the AI-102 objectives: still useful for the underlying Azure AI service concepts, but they are not an AI-103 preparation set.
Exam at a glance
Microsoft's associate-tier certification for AI engineers building production AI solutions on Azure.
Skills measured
- Plan and manage an Azure AI solution — 15–20%
- Implement generative AI solutions — 15–20%
- Implement an agentic solution — 10–15%
- Implement computer vision solutions — 15–20%
- Implement natural language processing solutions — 15–20%
- Implement knowledge mining and document intelligence — 10–15%
Core services tested
- Azure AI Foundry — portal + SDK for managing AI projects, model catalog, prompt flow, evaluations.
- Azure OpenAI Service — model deployment, embeddings, content filtering, RBAC, networking, prompt engineering patterns.
- Azure AI Vision & Custom Vision — image analysis, object detection, OCR, custom image classification + detection models.
- Azure AI Document Intelligence — prebuilt and custom models for layout, invoices, receipts, IDs.
- Azure AI Language — entity recognition, sentiment, custom text classification, custom NER, conversational language understanding.
- Azure AI Speech — speech-to-text, text-to-speech, custom neural voice, speaker recognition.
- Azure AI Search — indexes, indexers, semantic ranking, vector search for RAG patterns.
- Azure AI Content Safety — content filtering, prompt shields, groundedness detection.
Prerequisites
No formal prerequisites. Microsoft recommends hands-on experience developing solutions in Python or C#, comfort calling REST APIs and SDKs, and a working understanding of responsible AI principles. Passing AI-900 first is strongly recommended for non-developers entering the AI space.
Why take this certification
- Most-requested Azure AI credential. AI-102 is the only Microsoft certification dedicated to building production AI applications on Azure, and it is increasingly listed as a requirement on cloud AI engineering job postings.
- Competitive salary. Azure AI engineers in the United States typically earn $130,000–$160,000 USD per year, with senior generative-AI specialists trending higher as enterprises scale Azure OpenAI deployments.
- Aligned with the generative-AI shift. The most recent AI-102 update added dedicated skill areas for generative AI and agentic solutions — reflecting how Azure customers actually build with AI Foundry, Azure OpenAI, and the Agent Service today.
Note: this certification retired on 30 June 2026 and can no longer be earned or renewed — Microsoft lists AI-103 as its replacement. Status on the official certification page.
What you'll learn in the AI-102 exam
AI-102 targets developers and AI engineers building production AI solutions on Azure. The exam emphasizes Azure AI Foundry, Azure OpenAI, custom vision and language solutions, and integrating AI capabilities into applications — with a strong responsible-AI thread running through every skill area.
Azure AI Foundry and project management
- Provision and configure Azure AI services (single-service vs multi-service resources, keys, endpoints, custom domains).
- Use the Azure AI Foundry portal and SDK to create projects, deploy models from the catalog, and run prompt flow evaluations.
- Secure AI workloads with managed identity, RBAC, private endpoints, customer-managed keys, and network isolation.
- Monitor cost, throughput, and latency with Azure Monitor, diagnostic logs, and quota dashboards.
Generative AI and agents
- Deploy Azure OpenAI models — chat, embeddings, image generation — and tune deployment SKUs for throughput.
- Apply prompt engineering, few-shot patterns, structured outputs, and function/tool calling.
- Build agents with the Agent Service in AI Foundry, including connecting tools and using MCP for tool integration.
- Apply content filtering, prompt shields, and groundedness detection via Azure AI Content Safety.
Computer vision
- Use Azure AI Vision for image analysis, OCR, and spatial analysis scenarios.
- Train and deploy Custom Vision classification and object-detection models.
- Implement Face API responsibly within Microsoft's gated-access guardrails.
- Extract structured data with Azure AI Document Intelligence (prebuilt + custom models).
Natural language processing
- Use Azure AI Language for entity recognition, key phrase extraction, sentiment, and PII detection.
- Build custom text classification and custom NER models.
- Translate text and speech with Azure AI Translator.
- Implement speech-to-text, text-to-speech, and custom neural voice with Azure AI Speech.
Knowledge mining and RAG
- Build Azure AI Search indexes with skillsets, indexers, and cognitive enrichment.
- Implement semantic ranking and vector search to power retrieval-augmented generation.
- Combine Azure AI Search with Azure OpenAI to ground LLM responses in enterprise data.
How the practice exams help
Each free question and every premium exam mirrors the scenario-style format Microsoft uses — workload constraints, multiple correct answers in some items, drag-and-drop sequences, and code snippets where you choose the right SDK call. Detailed explanations cover not just why the right answer is right but why the distractors are wrong, so you learn Azure's AI surface rather than memorizing answers.
How to prepare for the AI-102 exam
A successful AI-102 preparation strategy combines theoretical study with substantial hands-on time on Azure AI Foundry and Azure OpenAI. Recommended approach:
- Establish fundamentals (1 week). If you have not already passed AI-900, work through the AI-900 learning path first. It locks in vocabulary (responsible AI, common AI workloads, Azure AI service families) that AI-102 assumes.
- Study the official AI-102 learning path (3–4 weeks). Review the official AI-102 study guide and work through the modules on Microsoft Learn — the official content closely mirrors the exam blueprint.
- Hands-on with Azure AI Foundry and Azure OpenAI (2–3 weeks). Create an Azure free-tier subscription and build real solutions: deploy a gpt-4 model in Azure OpenAI, build a RAG pipeline with Azure AI Search, train a Custom Vision detector, fine-tune a custom NER model, and wire up an agent in AI Foundry. The exam tests workflow knowledge that only hands-on use produces.
- Practice exams (1–2 weeks). Take timed practice tests to identify weak skill areas. Microsoft's free practice assessment is a useful baseline; supplement with third-party question banks for breadth. Aim for consistent 80%+ scores before scheduling.
Recommended timeline
6–10 weeks of focused study (10–15 hours per week) for developers with prior Azure experience. Engineers new to Azure should allow 10–14 weeks and pass AI-900 first.
Official resources
Bookmark the official AI-102 certification page and download the linked study guide. Microsoft's free Learn training hosts the official AI Engineer learning path, and the Microsoft Learn exam sandbox lets you experience the actual exam interface before exam day.
Frequently asked questions
Are these actual exam questions?
No. All Nex Arc Azure AI-102 practice questions are original content based on the provider's published exam blueprint, official documentation, and real-world scenarios across every topic the exam covers. We do not offer brain dumps, leaked questions, or actual exam content. Using or distributing real exam questions violates certification provider agreements and can result in certification revocation.
How many questions are in the AI-102 exam?
Microsoft does not publish an exact question count for AI-102. Most candidates report 40-60 questions delivered in 100 minutes. Question types include multiple choice, multiple response, drag-and-drop, case studies, and short-code or active-screen scenario items. The exam may include interactive components.
What's the AI-102 passing score?
The passing score is 700 out of 1000. Microsoft uses a scaled scoring model where item difficulty is weighted, so the raw percentage of items you answered correctly is not the same as your scaled score. Focus on understanding the underlying Azure AI services rather than memorizing answers.
How do I access the full AI-102 practice exam?
AI-102 retired on 30 June 2026, so the course is no longer sold on its own. The complete Azure AI-102 practice exams and study guide are included in Nex Arc Pro, which unlocks every active and deprecated course. Anyone who bought the AI-102 course before it retired keeps lifetime access — nothing changes for existing owners.
What are the prerequisites for the AI-102 exam?
There are no formal prerequisites for AI-102, but Microsoft recommends hands-on experience developing solutions in Python or C#, plus the ability to use REST APIs and SDKs to build AI solutions on Azure. Most candidates benefit from passing AI-900 (Azure AI Fundamentals) first to establish a baseline understanding of Azure AI services and responsible AI principles. Real-world experience with Azure AI Foundry, Azure OpenAI, Azure AI Search, and at least one of the AI service families (vision, language, speech) is strongly recommended.
How long is the AI-102 certification valid?
Microsoft role-based certifications, including AI-102, are valid for one year from the date you pass the exam. Renewal was free and was completed through an online assessment on Microsoft Learn. The certification and its renewal assessment retired on 30 June 2026: a credential earned before that date stays valid on your transcript until it expires, but it can no longer be renewed and the exam can no longer be taken.
What's the exam cost and can I retake it if I fail?
The AI-102 exam costs $165 USD in the United States (prices vary by country and are converted to local currency at booking). The exam is proctored and delivered through Pearson VUE, either at a testing center or via online proctoring. If you fail your first attempt, you must wait 24 hours before retaking. Subsequent failed attempts have progressively longer waiting periods (14 days, then 14 days, with a maximum of five attempts in a 12-month period). Each attempt requires a new exam fee. Microsoft occasionally offers Exam Replay vouchers that bundle a retake at a discount.
What Azure AI services are most heavily tested on AI-102?
Based on the six skill areas, the most frequently tested services are: Azure AI Foundry (portal and SDK for managing AI projects), Azure OpenAI Service (model deployment, prompt flow, embeddings, content filtering), Azure AI Vision and Custom Vision (image analysis, object detection), Azure AI Document Intelligence (form recognition, layout extraction), Azure AI Language (entity recognition, sentiment, custom text classification, custom NER), Azure AI Speech (speech-to-text, text-to-speech, custom voice), Azure AI Search (semantic ranking, vector search for RAG), and Azure AI Content Safety (prompt shields, groundedness detection). Together these account for the bulk of the exam content.
How does AI-102 differ from AI-900?
AI-900 is the fundamentals exam — it tests conceptual understanding of AI workloads, Azure AI services, and responsible AI principles. It is multiple choice only and assumes no prior coding experience. AI-102 is the associate-level engineering exam — it tests your ability to actually build, deploy, secure, and monitor production AI solutions on Azure using SDKs and REST APIs. AI-102 includes hands-on scenarios, case studies, and code-based questions. Most candidates pass AI-900 in 2-4 weeks of study; AI-102 typically requires 6-10 weeks for developers with prior Azure experience.
Can I retake the AI-102 exam if I fail?
Yes. Microsoft allows a retake 24 hours after your first failure. Subsequent retakes within the same 12-month period require longer waits, and you are capped at five attempts per 12-month rolling window. You pay the full exam fee for each attempt unless you purchased an Exam Replay voucher.
AI-102 retired on 30 June 2026 — what now?
Microsoft retired AI-102 and the Azure AI Engineer Associate certification on 30 June 2026, and lists Azure AI Apps and Agents Developer Associate (Exam AI-103) as its replacement. Anyone who already earned the certification keeps it on their transcript until it expires; it can no longer be renewed. New candidates take AI-103 instead. AI-103 is a different exam rather than a reissue of AI-102 — it centres on Microsoft Foundry, generative apps and production agents — so check its own skills-measured list on Microsoft Learn rather than assuming AI-102 preparation carries over.