The short version
Pick the 4-week plan if you already know some Azure and can give about 6 to 8 hours a week. Pick the 8-week plan if AI, Azure, or Python are new, at about 3 to 5 hours a week. These ranges are estimates. Your own pace matters more, so adjust after the first baseline practice set.
Both plans follow the two domains in Microsoft’s “Skills measured as of April 15, 2026”. Because 55-60% of the exam is about implementing with Microsoft Foundry, at least half of your time should be hands-on. Reading about a prompt is not the same as writing one and seeing what the model does.
Before you start
- Open the AI-901 study guide and keep it as your checklist.
- Create a free Azure account if you can, so you can open the Foundry portal and deploy a model. Watch for costs and delete resources when you finish. Our certification costs guide explains what to expect.
- Take one practice set cold, before studying. A low score here is useful, not discouraging. It shows where to start.
The 4-week plan
| Week | Focus | Hours |
|---|---|---|
| 1 | Domain 1: responsible AI, how generative models work, model choice, deployment options | 6-8 |
| 2 | Domain 2, part 1: prompts, deploying a model, chat client with the SDK, a single agent | 6-8 |
| 3 | Domain 2, part 2: text analysis, speech, vision and image generation, Content Understanding | 6-8 |
| 4 | Gaps, timed practice, review | 5-7 |
Week 1. Learn the six responsible AI principles until you can match a scenario to the right one in seconds. Then cover tokens, embeddings, and context windows, and the difference between a small and a large model. Learn the main deployment choices and the parameters you can change, such as temperature. Finish by listing the AI workloads: text analysis, speech, vision, and information extraction.
Week 2. Move into the portal. Deploy a model, write a system prompt and a user prompt, and compare the outputs. Read and run a short chat client built with the Foundry SDK. Create a single agent, give it a tool, and test it in the portal. Write down what each part of the code does.
Week 3. Work through the services one at a time. Run text analysis on a few sentences. Try speech to text and text to speech. Send an image to a multimodal model, and try image generation. Use Content Understanding on a document, an image, and an audio file if you can.
Week 4. Take a practice set, review every miss, and read the matching part of the study guide again. Repeat with a second set a few days later. Do not learn new material in the last two days.
The 8-week plan
| Week | Focus | Hours |
|---|---|---|
| 1 | Cloud and AI basics; set up an Azure account; Python reading practice | 3-5 |
| 2 | Responsible AI principles and how generative AI works | 3-5 |
| 3 | Choosing models, deployment options, parameters | 3-5 |
| 4 | Prompts, the Foundry portal, a chat client with the SDK | 3-5 |
| 5 | Agents, then text analysis | 3-5 |
| 6 | Speech, vision, and image generation | 3-5 |
| 7 | Content Understanding and information extraction; first full review | 3-5 |
| 8 | Timed practice, gap fixing, rest | 3-5 |
The first week is deliberately gentle. If you have never read Python, spend it on short examples: variables, functions, calling an API, and reading JSON. You do not need to become a developer. You need to read a ten-line snippet and say what it does.
From week 2 onward, end each week with a short self-quiz of ten to fifteen questions on that week’s topic. In week 7, take a full practice set and review it slowly. In week 8, take a second one under timed conditions.
How to use practice exams in the plan
- Once at the start, as a baseline, with no studying first.
- Once per topic block, untimed. Read every explanation, including for correct answers.
- Twice near the end, timed. Our five sets rise in difficulty, from single-concept questions to a full-length timed simulation.
- Track by domain, not only by total. A good total can hide a weak area. Practice exams show readiness but do not replace Microsoft Learn or hands-on time. See our how-to-study guide for general technique.
The last 3 days
- Three days out: one timed practice set. Review only the misses and write one line for each cause.
- Two days out: reread the responsible AI principles, the deployment and parameter notes, and your agent and Content Understanding notes. Try the exam sandbox so the interface is familiar.
- The day before: light review for an hour at most. Check your ID, your appointment, and the room or route. Sleep.
Our exam day guide has a checklist.
More guides for AI-901
- AI-901 Exam Format, Questions and Scoring
- AI-901 Sample Questions With Answers
- Is AI-901 Hard? An Honest Difficulty Check
- AI-901 Prerequisites: What to Know First
- AI-901 Renewal and Retakes Explained
- Is AI-901 Worth It? An Honest Look
- AI-901 Cheat Sheet: Key Facts to Review
- AZ-900 vs AI-901: Which First?
- AB-900 vs AI-901: Which Exam First?
- AWS AIF-C01 vs AI-901: Which First?
Next step
Use the domains explained page as your topic checklist. When you are ready for a timed run, try our AI-901 practice tests.
Frequently asked questions
Can I do it in two weeks?
Possibly, if you already work with Azure AI and Python. Compress the 4-week plan and keep the hands-on parts.
Do I need to pay for Azure?
Not necessarily. Many exercises are small, but costs depend on what you deploy. Check Microsoft’s pricing page and delete resources afterward.
What if my baseline score is very low?
That is common for people new to the topic. Use the 8-week plan and lean on Microsoft Learn first.