Exam overview — AI-103
Exam AI-103: Developing AI Apps and Agents on Azure leads to the Microsoft Certified: Azure AI Apps and Agents Developer Associate credential. It targets Azure AI engineers who build, manage, and deploy generative AI and agents on Microsoft Foundry, using Python and Azure AI services.
At a glance
| Credential | Azure AI Apps and Agents Developer Associate |
| Pass score | 700 / 1000 (scaled) |
| Length | ~120 minutes |
| Skills measured as of | 16 April 2026 |
| Question styles | Multiple choice, multiple response, case studies |
| Renewal | Annually, via a free Microsoft Learn assessment |
Skills measured — the five domains
| # | Domain | Weight |
|---|---|---|
| 1 | Plan and manage an Azure AI solution | 25–30% |
| 2 | Implement generative AI and agentic solutions | 30–35% |
| 3 | Implement computer vision solutions | 10–15% |
| 4 | Implement text analysis solutions | 10–15% |
| 5 | Implement information extraction solutions | 10–15% |
Domains 1 and 2 together are 55–65% of the exam — that's where Foundry models/agents, RAG, and orchestration live, so weight your prep accordingly.
What you should already be proficient in
- Python application development with the Azure and Foundry SDKs
- Generative AI concepts: prompting, RAG, embeddings, vector/hybrid/semantic search
- Microsoft Foundry — projects, model & agent deployments, Foundry Tools
- Agentic patterns — tools/function-calling, memory, multi-agent orchestration, approval flows
- Responsible AI — content filters, evaluators, safety, guardrails, and observability
Key cross-cutting themes
- Foundry-first — most tasks route through a Foundry project, its model/agent deployments, and connected tools.
- Grounding & retrieval — RAG with Azure AI Search (vector + hybrid + semantic) recurs across domains 2 and 5.
- Agents — roles/goals, tool schemas, memory, function-calling, and multi-agent orchestration.
- Responsible AI & observability — content filtering, evaluators, tracing, and safety signals everywhere.
- Keyless security — managed identity, RBAC, and private networking over stored keys.
Ready? Start with Domain 1 or jump to the study plan.
Content is grounded in the official Microsoft Learn AI-103 study guide. Always confirm details against current Microsoft documentation.