Microsoft Azure AI Fundamentals (Exam AI-900) is the no-code, one-exam way to prove you understand AI, machine learning, and generative AI on Azure. Practice with real-style questions and walk in knowing exactly what's tested.
Be clear-eyed about what this is: AI-900 is an entry-level, fundamentals certification. It validates that you understand AI concepts and how Microsoft's AI services fit together. It does not make you a machine learning engineer, and on its own it rarely lands a senior AI job. If you want a credential that signals deep, hands-on engineering skill, look at the associate-level Azure AI Engineer track instead.
That said, AI-900 earns its place for the right people. It is genuinely useful if you are:
A career-changer or student getting your first AI credential and want a structured starting point. A non-technical professional (sales, product, marketing, project management, operations) who needs to talk credibly about AI capabilities with engineering teams and customers. A developer or IT pro new to Azure AI who wants a fast, mapped overview before going deeper. Someone building an Azure certification path who wants an easy, confidence-building win before tackling associate-level exams.
The best part for newcomers: no coding is required and there are no prerequisites. Microsoft states that data science and software engineering experience are not needed, though basic familiarity with cloud concepts and client-server applications helps. Most candidates can prepare in a few focused weeks. For a low-cost, low-risk way to demonstrate AI literacy on your resume, it does the job. If you expect it to do more than that, you will be disappointed.
The exam is built around five skill areas. The percentages below are the official weightings from Microsoft's published skills-measured outline, so they tell you where to spend your study time. Note that generative AI is now the single largest area, so do not treat it as an afterthought.
Features of generative AI models and common scenarios, responsible AI considerations for generative AI, and Azure services including Azure OpenAI Service and Azure AI Foundry (including its model catalog).
Identifying common AI workloads (computer vision, NLP, document processing, generative AI) and the six guiding principles of responsible AI: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability.
Core ML techniques (regression, classification, clustering, deep learning, the Transformer architecture), features versus labels, training and validation datasets, and Azure Machine Learning capabilities like automated ML and model deployment.
Image classification, object detection, optical character recognition (OCR), and facial detection and analysis, plus the Azure AI Vision and Azure AI Face services.
Key phrase extraction, entity recognition, sentiment analysis, language modeling, speech recognition and synthesis, and translation, delivered through the Azure AI Language and Azure AI Speech services.
Question formats include multiple choice and drag-and-drop. You'll be asked to recognize the right Azure service for a scenario far more often than to recall a definition, so understanding what each service does beats rote memorization.
| Exam Code | AI-900 |
| Certification | Microsoft Certified: Azure AI Fundamentals |
| Cost | $99 USD (Microsoft uses regional pricing, so your local price may differ) |
| Time Limit | 60 minutes (plus check-in time) |
| Passing Score | 700 out of 1000 (scaled, not a straight percentage) |
| Question Formats | Multiple choice and drag-and-drop |
| Prerequisites | None; no coding or data science experience required |
| Vendor | Microsoft |
Microsoft does not publish a fixed question count for AI-900, and the number can vary; plan your pacing around the 60-minute clock rather than a set number of items. Confirm all details on the official Microsoft page before booking.
You don't need months. A focused two-to-four week plan works for most people. Here's a sequence that maps directly to how the exam is weighted:
Open Microsoft's AI-900 study guide and treat its bullet list as your checklist. Every exam question maps back to one of those objectives.
Generative AI is the heaviest area (20–25%) and responsible AI's six principles show up repeatedly. Learn these cold before anything else.
Build a one-page table: for each Azure service (AI Vision, AI Face, AI Language, AI Speech, Azure OpenAI, Azure AI Foundry, Azure Machine Learning), write what problem it solves. The exam rewards this.
You can pass without it, but 30 minutes clicking through services in the Azure portal or AI Foundry makes the concepts stick far better than reading alone.
Once you've read the material, switch to active recall. Practice questions expose the gaps you didn't know you had and train you for the scenario-style wording.
Microsoft offers a free practice assessment. Use it near the end to gauge readiness and find weak domains before exam day.
Reading a study guide tells you what you've seen. Practice questions tell you what you actually know. That gap is where people fail fundamentals exams: they recognize a topic on the page, assume they've got it, and then freeze when the exam reframes it as a scenario ("Which Azure service should a team use to extract text from scanned invoices?").
Active recall and testing are among the most reliable ways to move information into long-term memory. Working questions also trains the skill the AI-900 actually measures: picking the right service for a situation under a time limit. And every question you get wrong is a precise, free pointer to exactly what to review next.
GetMyCert's AI-900 questions come with detailed explanations, not just an answer key, so you learn why the right choice is right and why the distractors are wrong. That reasoning is what carries over to the real exam.
Always verify current details and book your exam through Microsoft directly:
You need a scaled score of 700 out of 1000 to pass. Because it's scaled, 700 does not mean you answered exactly 70% of questions correctly; harder items are weighted differently.
The standard price in the United States is $99 USD. Microsoft uses regional pricing, so your local fee may be higher or lower. Confirm the exact price for your country on the official Microsoft certification page.
You get 60 minutes to complete the exam, with additional time allotted for check-in. If the exam isn't offered in your preferred language, you may be able to request extra time.
No. Microsoft states that data science and software engineering experience are not required. The exam is designed for both technical and non-technical candidates. Basic awareness of cloud concepts and client-server applications is helpful but not mandatory.
Five skill areas: AI workloads and responsible AI considerations, machine learning principles on Azure, computer vision, natural language processing, and generative AI workloads. Generative AI carries the highest weight at 20–25%.
It's a solid entry-level credential for career-changers, students, and non-technical professionals who need to discuss AI credibly, or for anyone starting an Azure certification path. It is not a substitute for deeper, associate-level certifications if you want to work as an AI engineer.
No. There are no required prior exams or qualifications. You can take AI-900 as your very first Microsoft certification.
Most candidates are ready in about two to four weeks of focused study, depending on their background. If you already work with cloud or AI tools, you may need less; if AI is entirely new, give yourself the full window and lean on practice questions.
Work through original AI-900 practice questions with full explanations and find your weak spots before exam day.
Start Practicing Free