AI-901

AI-901: Microsoft Azure AI Fundamentals

Everything you need for Microsoft exam AI-901: format, passing score, the two skill areas, a study plan and free sample questions.

Level
Fundamentals
Passing score
700 / 1000
Replaces
AI-900
Hands-on
55–60%

AI-901 is Microsoft’s entry-level exam for people starting a career in AI solution development on Azure. Passing it earns the Microsoft Certified: Azure AI Fundamentals credential. It replaced AI-900, which Microsoft retired on June 30, 2026.

Do not let the word “fundamentals” mislead you. This is no longer a purely conceptual exam. More than half the score sits in a skill area called implement AI solutions by using Microsoft Foundry, and Microsoft’s own audience profile now expects knowledge of Python syntax and familiarity with REST APIs, SDKs and CLIs. AI-900 asked you to describe things; AI-901 asks you to build small things.

What the exam covers

Skill areaWeight
Identify AI concepts and capabilities40–45%
Implement AI solutions by using Microsoft Foundry55–60%

Two areas only, which makes this a simpler exam to plan for than most. The first covers responsible AI, how generative models work, choosing a model, and recognising which AI workload fits a scenario. The second is hands-on: prompts, deploying a model, a small chat client, a single agent, text and speech, vision, and information extraction with Content Understanding.

Where to start

New to all of this? Read the exam format guide first, then the prerequisites to check you have the background Microsoft assumes. Coming from AI-900? Start with what changed — it is more than a renumbering. If you would rather learn the material step by step, the free written course covers both skill areas in 12 lessons with diagrams.

AI-901 guides

  1. AI-901 free course: Azure AI FundamentalsA free written AI-901 course in 12 lessons with diagrams: AI concepts, responsible AI and building with Microsoft Foundry, in exam order.
  2. AI-901 exam format: questions, passing score, time and costThe AI-901 format: passing score, languages, how it is delivered, what is published about length and price, and the question types to expect.
  3. AI-901 vs AI-900: what changed?The differences between AI-901 and the retired AI-900: two domains instead of five, hands-on Foundry work, and no more classic machine learning.
  4. AI-901 study plan: pass in 4 weeksA four-week AI-901 study plan weighted by the two official skill areas, with hands-on Foundry tasks for each week.
  5. AI-901 sample questions: 5 worked examplesFive worked AI-901 sample questions across both skill areas, each with an explained answer.
  6. AI-901 identify AI concepts and capabilities explainedThe AI-901 concepts domain explained: the six responsible AI principles, model choice and configuration, and recognising AI workloads.
  7. AI-901 implement AI solutions with Foundry explainedThe largest AI-901 domain explained: prompts, model and agent deployment, the Foundry SDK, speech, vision and Content Understanding.
  8. Is AI-901 hard? An honest look at the difficultyHow hard AI-901 really is, who finds it easy, who struggles, and the specific things that catch people out on exam day.
  9. AI-901 exam tips: how to gain marks on the dayPractical AI-901 exam tips: how to read the question types, the distinctions that decide marks, and how to manage the sitting itself.
  10. AI-901 prerequisites: what you need before you startThere are no formal AI-901 prerequisites, but Microsoft assumes Python, Azure familiarity and API basics. Here is what that means in practice.
  11. Is AI-901 worth it? Who should take the examIs AI-901 worth taking? Who benefits, what it costs, and when a role-based Azure certification is the better use of your time.
  12. Free AI-901 practice questions: 20 with answers20 free AI-901 practice questions covering AI concepts and Microsoft Foundry, with an explanation for every answer.
  13. AI-901 study resourcesThe free Microsoft material that covers AI-901 properly, why hands-on time still helps, and what not to bother with.
  14. AI-901 cheat sheetA one-page AI-901 reference: the two skill areas, the capability distinctions tested, and the over-preparation trap.
  15. AI-901 career: what the credential leads toMicrosoft tags Azure AI Fundamentals with one job role, AI Engineer. What AI-901 shows, what it leaves unproven, and which exams realistically follow.
  16. AI-901 renewal, validity and retake rulesAI-901 never expires, but retakes follow strict waits: 24 hours, then 14 days, max five attempts a year. Plus what AI-900's retirement means for holders.
  17. AI-901 study scenarios: three profilesPlan AI-901 as a non-coder, a developer new to AI or an AI-900 holder: why implementation is 55 to 60% of the score and where each profile loses time.