In the full course
What you get
- 400 exam-style questions, each with a full explanation
- 200 flashcards, filtered by domain
- The full study guide, 27 chapters
- Timed mock exams matched to the real exam length
- A readiness score weighted by the official exam blueprint
Get full access to Microsoft AI-103
All questions, timed exams, flashcards, PDF study guide download & progress tracking.
Lifetime · all courses
$29.99
One payment · future courses included
30 seconds, then straight to checkout.
Pass, or your money back
Reach 85% readiness on this course, sit the real exam, and if you don't pass we refund it in full. Applies to this single-course purchase. Terms.
More free samples
Marked, and passed
Real feedback from people who passed
“I failed my CISSP on the first attempt with another platform. Switched to CertCrush, focused on my weak domains using the tracking feature, and passed three months later. The explanations for wrong answers are genuinely useful, not just 'A is correct because A is correct'.”
“Honestly wasn't expecting much but this is probably the best ten bucks I've spent on exam prep. Did 20–30 questions every morning before work for 6 weeks. Passed with a comfortable margin. The timed exam mode is what really got me comfortable with the pressure.”
“The flashcards are underrated. I used them during my commute and it made a huge difference for the theory-heavy ITIL questions. Passed first try. Already using it again for CISM.”
Microsoft AI-103 replaced AI-102 as the current Azure AI Apps and Agents Developer Associate exam. It tests Python-based development of generative AI applications and autonomous agents using Microsoft Foundry, alongside computer vision, text analysis, and information extraction. This course covers all five exam domains with practice questions weighted to the official blueprint.
Practice content last updated · Independently written and aligned to Microsoft’s published exam objectives.
About the Microsoft AI-103 Exam
AI-103 is Microsoft's current certification for engineers who build generative AI applications and autonomous agents on Azure. It replaced AI-102 in mid-2026 as Microsoft shifted its AI certification track toward agentic development, and the content reflects that shift: nearly a third of the exam covers building and orchestrating agents with Microsoft Foundry, not just calling a single model endpoint. Candidates coming from AI-102 or general Azure experience often underestimate how much of AI-103 sits outside plain prompt-response work. The exam expects you to design multi-agent orchestration patterns, implement retrieval-augmented generation pipelines with proper grounding, and reason about agent memory and tool-calling schemas, not just describe what RAG is. A second common gap is responsible AI: the exam tests concrete implementation, such as configuring content filters, detecting prompt injection through embedded image text, and setting up trace logging for agent auditing, rather than general principles. The exam also spans computer vision, text analysis, and information extraction, so a candidate who is strong on generative AI but has not touched Content Understanding pipelines or Azure AI Search's hybrid and vector search options will lose points outside their comfort zone. This course maps every practice question to one of the five official exam domains at their published weighting, so time spent studying tracks time spent being tested. Explanations point to the Foundry service or SDK pattern behind each answer, not just which option is correct.
Exam Domains Covered
- Plan and manage an Azure AI solution28%
- Implement generative AI and agentic solutions33%
- Implement computer vision solutions13%
- Implement text analysis solutions13%
- Implement information extraction solutions13%
Exam Format & Details
50 questions in a 100-minute associate exam (standard Microsoft associate format; Microsoft does not publish question types or lab presence in advance for any exam). Passing score is 700 out of 1000, which is 70%. Voucher price is $165 USD, standard Associate/Expert pricing. Book through Pearson VUE from your Microsoft Learn profile.
Why Practice Questions Matter
AI-103 tests implementation details that are easy to skim past in documentation: which Foundry service handles vector search versus hybrid search, how agent memory and tool schemas fit together, and where responsible AI controls actually get configured versus just described. Practice questions surface these gaps before exam day, when guessing costs a retake fee and another 100 minutes. Working through domain-weighted questions also corrects a common study mistake: spending equal time across all five domains when generative AI and agentic solutions alone make up roughly a third of the exam.
Sample Practice Questions
The following questions are a preview of the type of syllabus-aligned questions you will practise in CertCrush. They reflect the format and reasoning style of the Microsoft AI-103 exam, not actual exam content.
Q1.Which types of images should be scanned through Azure AI Content Safety's image analysis in a responsible visual content pipeline?
- A.Both content people upload or submit and content the system itself generates
- B.Only content that users upload, since generated content is inherently safe
- C.Only content the system generates, since user uploads cannot violate policy
- D.Only images larger than a certain file size
Domain: Implement computer vision solutions
Q2.A global standard deployment routes requests:
- A.To whatever datacenter in Microsoft's global capacity pool has availability
- B.Only to a single fixed region chosen at deployment time
- C.Exclusively to on-premises infrastructure
- D.Only within one country regardless of load
Domain: Plan and manage an Azure AI solution
Q3.Groundedness detection, as a risk and safety feature, flags:
- A.Deployments that are missing a managed identity
- B.Requests that exceed a rate limit
- C.Responses that contain profanity
- D.Responses that make claims unsupported by the provided source material
Domain: Plan and manage an Azure AI solution
Q4.Why does hosting a sensitive tool as a separately-secured Azure Function, rather than as a simple in-process function, offer a governance advantage?
- A.It allows tool access and the underlying resource's access to be governed independently
- B.It automatically increases the model's context window
- C.It removes the need for a tool schema entirely
- D.It is required before any function calling can work at all
Domain: Implement generative AI and agentic solutions
Q5.A team notices their generative app consistently answers in the wrong tone and mishandles a particular category of ambiguous request, even though the retrieved context looks accurate and complete every time. What is the most likely layer to address first?
- A.Prompt engineering or fine-tuning, not the retrieval pipeline
- B.Rebuilding the vector index from scratch
- C.Switching to a different search connector type
- D.Increasing the number of chunks retrieved per query
Domain: Implement generative AI and agentic solutions
Q6.An agent is misfiring on tool selection, frequently calling the wrong tool for a given request out of a toolkit of a dozen loosely-defined functions. What is the most likely fix, according to standard exam-tested guidance?
- A.Narrow and more precisely describe the tool schema, rather than adding more tools
- B.Add several additional tools to give the agent more options
- C.Increase the model's temperature parameter to improve tool selection
- D.Remove the system prompt entirely to reduce ambiguity
Domain: Implement generative AI and agentic solutions
Q7.A team asks a model to 'walk through your calculation step by step, then give the total cost' instead of simply asking for the total cost directly. What technique does this represent, and what is its main benefit?
- A.Chain-of-thought evaluation, which exposes the reasoning trail for debugging and for tasks needing visible logic
- B.Self-critique looping, which requires a completely separate evaluator model
- C.Groundedness scoring, which only applies to retrieval-augmented responses
- D.Memory summarization, which compresses older parts of a conversation
Domain: Implement generative AI and agentic solutions
Q8.For provisioned throughput deployments, capacity is expressed in:
- A.Azure RBAC role assignments
- B.Requests per minute only
- C.Provisioned throughput units (PTUs)
- D.Tokens per minute only
Domain: Plan and manage an Azure AI solution
Q9.A team designs a single Foundry tool that can search a product catalog, check live inventory, and place orders, all within one tool definition, because it seemed simpler to build. What is the main planning-stage risk with this design?
- A.Foundry technically prevents a single tool from performing more than one action
- B.This design eliminates the need for any retrieval strategy
- C.A do-everything tool is harder for the model to reason about correctly and harder to secure and audit later than several narrowly scoped tools
- D.Combining capabilities into one tool always increases latency dramatically
Domain: Plan and manage an Azure AI solution
Q10.Which statement correctly distinguishes the chunking decision from the embedding model decision when building the retrieval half of a RAG pipeline?
- A.Chunking controls how documents are split, while the embedding model controls how each piece is converted into a meaning-representing vector
- B.Chunking and embedding model selection are the same configuration setting
- C.The embedding model only matters for multimodal content
- D.Chunking has no effect on retrieval quality once embeddings are generated
Domain: Implement generative AI and agentic solutions
Microsoft AI-103 guides & exam news
AI-102 Is Retired: What AI-103 Changes and Which to Sit Now
AI-102 retires on 30 June 2026 and AI-103 is its successor. Here is a clear, side-by-side comparison of the two Azure AI certifications, plus a straight answer on which one you should take this year.
How to Pass the Microsoft AI-103 Exam in 2026: An 8-Week Study Plan for the New Azure AI App and Agent Developer Certification
Microsoft's AI-102 retires on 30 June 2026 and the new AI-103 Azure AI Apps and Agents Developer exam takes its place. Here is a realistic 8-week study plan covering every domain, the exam format, the cost and the agent and RAG skills you need to pass first time.
SC-500 vs SC-200 in 2026: Which Microsoft Security Exam Should You Take?
SC-500 is for engineers who build and harden Azure, hybrid and AI workloads. SC-200 is for analysts who detect, investigate and respond. How the two 2026 exams differ, where their objectives overlap, and which to take first.
Microsoft AB-650 Exam Topics and Domain Weightings in 2026: The MS-102 Replacement, Now in Beta
AB-650 replaces MS-102 as Microsoft's tenant administration exam, and the weightings tell a different story. Security and governance take 40 to 45 percent, AI services take 35 to 40 percent, and classic tenant admin is down to a fifth of the exam.
Frequently Asked Questions
What is included in the free Microsoft AI-103 sample?
The free sample includes 10 syllabus-aligned practice questions, sample flashcards, and a preview chapter from the study guide. No account or payment is required to try the sample.
How many questions are in the full Microsoft AI-103 course?
The full Microsoft AI-103 course includes 400 practice questions, covering all 5 exam domains. Every question carries a full explanation for the right answer and the wrong ones.
Are these official Microsoft exam questions?
No. CertCrush questions are independently written and syllabus-aligned. They mirror the format, difficulty, and reasoning style of the official exam. We are not affiliated with or endorsed by Microsoft.
Which domains does the Microsoft AI-103 course cover?
The course covers 5 exam domains: Plan and manage an Azure AI solution, Implement generative AI and agentic solutions, Implement computer vision solutions, Implement text analysis solutions, Implement information extraction solutions.
Can I study on mobile?
Yes. CertCrush is fully responsive and works on phones, tablets, and desktops. The timed exam, flashcards, and study guide all work on mobile without installing an app.
What happens when I create an account?
Creating a free account lets you access full courses, track your weak areas by domain, and resume practice sessions across devices. No credit card is required to register.
Start with 10 free questions
No account, no card. The full Microsoft AI-103 course is $9.99, once.