The short version
AI-103 is Microsoft’s associate-level exam for people who build AI apps and agents on Azure with Microsoft Foundry. You need Python experience and a working feel for generative AI, not just a list of service names. This guide follows the study guide Microsoft lists as “Skills measured as of April 16, 2026”.
What the exam is
The full exam name is Developing AI Apps and Agents on Azure. Passing it earns the Microsoft Certified: Azure AI Apps and Agents Developer Associate credential, which Microsoft labels as intermediate level. The credential covers designing, developing, and deploying Azure AI solutions with Python and Microsoft Foundry.
Microsoft describes the person who takes this exam as an Azure AI engineer who builds, manages, and deploys agents and AI solutions. That person plans and manages AI solutions, builds generative and agentic solutions, and implements vision, text analysis, and information extraction. In practice, expect questions that start with a small scenario (“a team needs…”) and ask which service, setting, or design fits.
A note on history: the older AI-102 exam and its Azure AI Engineer Associate certification are retired, according to Microsoft’s own certification page. AI-103 is the current exam in this area, and its objectives lean toward Foundry, agents, and grounding rather than the classic standalone AI services.
Who it is for
Microsoft says you should have experience developing apps with Python, and you should be familiar with general AI, generative AI, and Azure services. You will work with business stakeholders, solution architects, data scientists, DevOps engineers, and cloud security engineers, so the exam also touches on security, monitoring, and governance. The prerequisites page lists what to know before you start.
Format at a glance
| Item | What Microsoft says |
|---|---|
| Exam code | AI-103 |
| Level | Intermediate |
| Time | 120 minutes |
| Passing score | 700 or higher on a scale of 1 to 1,000 |
| Delivery | Proctored, scheduled through Pearson VUE |
| Languages | 10, including English, Spanish, French, and Portuguese (Brazil) |
| Number of questions | Not published for this exam; Microsoft says most exams have 40 to 60 |
| Extras | May include interactive components |
Details on scoring, breaks, and retakes are on the exam format page.
The five domains and their weightings
| Domain | Weighting |
|---|---|
| Plan and manage an Azure AI solution | 25-30% |
| Implement generative AI and agentic solutions | 30-35% |
| Implement computer vision solutions | 10-15% |
| Implement text analysis solutions | 10-15% |
| Implement information extraction solutions | 10-15% |
Two domains carry well over half the weight, so they deserve well over half your time.
Plan and manage an Azure AI solution
Choosing models and Foundry services, setting up deployments and infrastructure, managing quotas and cost, monitoring, security (managed identity, private networking, keyless credentials), and responsible AI controls such as safety filters, evaluators, and audit trails.
Implement generative AI and agentic solutions
The largest domain. It covers building generative apps with retrieval-augmented generation, building agents with tools, memory, and multi-agent orchestration, and tuning and observing these systems.
Implement computer vision solutions
Generating and editing images and video, understanding images and video with multimodal models and Content Understanding, and keeping visual content safe.
Implement text analysis solutions
Extracting entities, summaries, and structured JSON with language models, detecting sentiment and sensitive content, translation, and speech to text and text to speech for agents.
Implement information extraction solutions
Building retrieval and grounding pipelines (ingestion, indexing, semantic, hybrid, and vector search) and extracting content from documents with OCR, layout analysis, and Content Understanding analyzers.
Official source
For the vendor’s own page, see the Microsoft Learn AI-103 study guide. It is the final authority if anything here differs, and Microsoft updates it from time to time.
More guides for AI-103
- AI-103 Study Plan: 4-Week and 8-Week Schedules
- AI-103 Sample Questions With Answers
- Is the AI-103 Exam Hard? An Honest Look
- AI-103 Renewal and Retakes Explained
- Is AI-103 Worth It? An Honest Look
- AI-103 Cheat Sheet: Key Facts to Review
- AI-103 vs AI-200: Which Exam Should You Take?
- AB-620 vs AI-103: Which Exam First?
Next step
Read the domains explained page to see where your time should go. When you want to check yourself, our AI-103 practice tests are original practice sets arranged around these five domains.
Frequently asked questions
Does AI-103 expire?
Yes. Microsoft says associate, expert, and specialty certifications expire annually. You renew by passing a free online assessment on Microsoft Learn during a six-month window before the expiry date.
Is AI-103 the same as AI-102?
No. AI-102 and its certification are retired according to Microsoft. AI-103 has different domains and focuses on Foundry, agents, and grounding.
Can I use Microsoft Learn during the exam?
Microsoft says Learn is available during role-based exams, with the exam clock still running. Details are on the exam format page.