AI-901 in 30 Days: A Week-by-Week Azure AI Fundamentals Study Plan (2026)

Last updated: July 2026. Researched against the official Microsoft Learn AI-901 exam page, skills outline, and certification policy pages.

TL;DR AI-901 replaced the retired AI-900 as Microsoft's Azure AI Fundamentals exam. It has two domains: AI concepts (40-45%) and hands-on implementation in Microsoft Foundry (55-60%). This 30-day plan splits the two free official learning paths across four weeks, weighting most of your time toward the practical Foundry work that now decides the exam.

Exam AI-901: Microsoft Azure AI Fundamentals is the certification that replaced AI-900 when Microsoft retired the older exam on 30 June 2026 [1][4]. Both exams award the same lifelong credential, Microsoft Certified: Azure AI Fundamentals, but AI-901 is a very different test. Where AI-900 asked you to describe what Azure AI services do, AI-901 asks you to build AI applications and agents in Microsoft Foundry [2][4]. That single shift, from describing to doing, is why a study plan matters more than it used to. You cannot pass AI-901 by reading alone.

This guide turns the official skills outline into a realistic four-week schedule. It assumes you can commit roughly one to one and a half hours a day, five to six days a week, which lines up with the amount of content in Microsoft's own instructor-led course for the exam [7]. If you have more time, compress it; if you have less, stretch the same blocks across six weeks without changing the order.

If you are still deciding between the old and new exams, read our companion piece on what changed between AI-900 and AI-901 first, then come back here to plan your preparation.

What AI-901 actually tests

Microsoft publishes the exam's skills outline on the official AI-901 study guide page, and it is refreshingly short [2]. As of the current version, "skills measured as of April 15, 2026", AI-901 has only two functional groups. For a broader overview before you plan, see our full AI-901 exam guide:

The headline planning fact is in the second row. The larger domain is entirely practical. It measures whether you have actually deployed a model, built a lightweight client application, created and tested an agent, and used Foundry Tools inside the Foundry portal and SDK [2]. Reading about those tasks is not the same as doing them, so your schedule needs to protect time at a keyboard.

Domain 1, Identify AI concepts and capabilities (40-45%), covers three areas: the principles of responsible AI (fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability); how generative AI models work and how to choose and configure them; and the common AI workloads, including generative and agentic AI, text analysis, speech, computer vision, and information extraction [2]. This is vocabulary and judgement rather than code, and it underpins everything in Domain 2.

Domain 2, Implement AI solutions by using Microsoft Foundry (55-60%), splits into four roughly equal hands-on blocks [2]:

• Generative AI apps and agents: write system and user prompts, deploy a model in the Foundry portal, build a lightweight chat client with the Foundry SDK, and create and test a single-agent solution.
• Text and speech: build a small text-analysis application and respond to spoken prompts using Azure Speech in Foundry Tools.
• Computer vision and image generation: interpret visual input with a multimodal model and generate new visual output.
• Information extraction: use Azure Content Understanding in Foundry Tools to pull structured data out of documents, images, audio, and video.

Microsoft does not publish a percentage split beneath the 55-60% headline, so treat those four blocks as four equal-sized study sessions.

Before you start: what AI-901 assumes you know

AI-901 is a Fundamentals-tier exam, but it is not a beginners-with-zero-experience exam in the way AI-900 was. The official audience profile states you should have "conceptual knowledge of AI solutions in Azure", plus "knowledge of Python coding syntax and programming techniques", and familiarity with "Azure resources", REST APIs, SDKs, and CLIs [1].

In plain terms, before day one you should be comfortable reading a short Python script, creating a resource in the Azure portal, and understanding what an API call is. If any of those is unfamiliar, add a few hours of Python basics and Azure portal orientation before Week 1. This is the single biggest audience change from AI-900, which assumed no coding background at all, so do not skip this honest self-check.

You do not need any prior certification. AI-901 is not gated behind another exam, and it sits below the role-based Associate and Expert credentials rather than requiring them [1].

The 30-day AI-901 study plan, week by week

The plan below maps the two free official Microsoft Learn paths, twelve modules in total, straight onto the two exam domains, then front-loads the concepts and gives the hands-on Foundry work three of the four weeks [2][8][9]. Microsoft gives no official day-by-day pacing, so treat this as a sensible, evidence-based structure rather than a rule.

Week 1, Domain 1: AI concepts (40-45%)

Goal: finish the conceptual domain in full and get your first look at the exam interface.

• Work through all six modules of the "AI concepts for developers and technology professionals" learning path: AI concepts, generative AI and agents, natural language processing, speech, computer vision, and AI-powered information extraction [8].
• Pay real attention to responsible AI. The six principles (fairness, reliability and safety, privacy and security, inclusiveness, transparency, accountability) are reliably examinable and easy marks if you learn them properly [2].
• End the week by trying the free exam sandbox, an interactive demo of the real exam interface and question styles, at aka.ms/examdemo. Getting familiar with the tooling early removes a source of exam-day stress.

By the end of Week 1 you should be able to match any given scenario to the correct workload (text, speech, vision, extraction, generative, or agentic) and explain why.

Week 2, Domain 2 part one: Foundry foundations, generative AI, and agents

Goal: get hands-on in Microsoft Foundry for the first time.

• Complete the first two modules of the "Get started with AI applications and agents on Azure" path: "Get started with AI in Azure" and "Get started with generative AI and agents in Azure" [9].
• Deploy your first model in the Foundry portal, then interact with it. Write a couple of system and user prompts and note how the output changes.
• Build the lightweight chat client with the Foundry SDK, and create and test a single-agent solution. These are exactly the tasks the outline names, so do them rather than read about them [2].

This is the week where the exam stops being theoretical. Budget extra time here if the Foundry portal is new to you.

Week 3, Domain 2 part two: text, speech, and computer vision

Goal: cover the middle two hands-on blocks of Domain 2.

• Work through "Get started with text analysis in Azure", which has you build a small Python client that analyses text, and "Get started with speech in Azure", which uses Azure Speech in Foundry Tools to recognise and synthesise speech [9].
• Then complete "Get started with computer vision in Azure", covering image analysis, image generation, and multimodal models [9].
• Keep a short notebook of the service names and what each one does. The exam rewards precise recognition of which Foundry Tool solves which problem.

Week 4, Domain 2 part three: information extraction, and consolidation

Goal: finish the syllabus and start revising.

• Complete "Get started with AI-powered information extraction in Azure", which uses Azure Content Understanding in Foundry Tools to extract structured information from documents, images, audio, and video [6][9]. Content Understanding is generally available and appears four times in the outline, so it is a first-class exam topic, not a footnote [2][6].
• Re-run any exercise from Weeks 1 to 3 that felt shaky. The hands-on tasks stick far better on a second pass.
• Read the skills outline one line at a time and tick off every bullet you can confidently perform or explain [2].

Final days: review and booking

• Repeat the exam sandbox until the interface feels routine.
• Do a full pass over the responsible AI principles and the workload-to-service mappings, your highest-yield revision.
• Book your Pearson VUE slot, either at a test centre or online through OnVUE [1][11]. If you are sitting it online, confirm now that you have a private room, a government-issued photo ID whose name matches your certification profile exactly, and a webcam for the room scan [11].

Practising under time pressure is the last piece. The timed tests on our AI-901 exam page help you rehearse the pace before the real thing.

How AI-901 differs from AI-900

If your notes or an older course still say "AI-900", here is what changed. AI-900 had five conceptual domains, every one beginning with the word "Describe" [4]. AI-901 has two, and the majority domain is hands-on [2]. The table below is the quickest way to see the gap.

Anyone who already passed AI-900 keeps that certification for life, and nothing is retroactively invalidated [4]. If you were mid-way through AI-900 preparation when it retired, our guide to the AI-900 retirement and what happens next explains exactly how to carry your progress across.

A 2026 currency check: names that changed

This is the part of AI-901 preparation most likely to trip you up, because Microsoft renamed the underlying platform. Study from anything published before 2026 and you will learn the wrong names.

• Microsoft Foundry is the current product name. It was previously called Azure AI Studio, then Azure AI Foundry [5]. The AI-901 outline uses the current name throughout, so quoting the outline is safe, but do not let older tutorials teach you the old one [2][5].
• Foundry Tools is the current name for what used to be called Azure AI Services [5].
• The agent API moved from the Assistants API to the Responses API, and the SDK story consolidated around a unified azure-ai-projects package [5].
• Azure Content Understanding now sits inside Foundry Tools and is generally available [6].

There is one more trap worth flagging. The AI-901 study guide's own "Find documentation" table still links to Language Understanding (LUIS) and Anomaly Detector, both carried over unedited from the AI-900 template [2]. LUIS was fully retired on 31 March 2026, and Anomaly Detector is scheduled to retire on 1 October 2026 [13]. Neither appears anywhere in the AI-901 skills outline, so do not follow those links expecting in-scope material. Prepare from the two current learning paths and the Foundry documentation instead [8][9].

Exam-day logistics

• Format and duration. AI-901 is a Fundamentals-tier exam with a 45-minute time limit and a 65-minute seat time; the extra minutes cover check-in and instructions [10]. Microsoft does not disclose an exact question count for AI-901, but states its certification exams generally contain 40 to 60 questions [10]. Fundamentals exams do not include performance-based labs and do not give in-exam access to Microsoft Learn, both of which are reserved for Associate and Expert exams [10].
• Passing score. You need 700 out of 1,000. This is a scaled score, not a percentage, so 700 does not mean 70% of the questions [3]. Multi-part questions can award partial credit, and there is no penalty for a wrong answer, so never leave a question blank.
• Fee. Microsoft prices exams by region and shows the figure at Pearson VUE checkout rather than on the exam page [1]. Fundamentals-tier exams like AI-901 are typically around US$99 in the United States [12]; check the live price for your own country when you register.
• Retakes. If you fail, you can retake after 24 hours; later retakes have longer waits, and there are no refunds for a failed or missed attempt [3].

Free official study resources

Everything in this plan can be done at no cost using Microsoft's own material:

• The two free learning paths, twelve modules in total, one path per domain [8][9].
• AI-901T00-A: Introduction to AI in Azure, the instructor-led course, roughly a day of content or about 24 hours self-paced [7].
• The exam sandbox at aka.ms/examdemo for the interface and question styles [1].

A free official practice assessment for AI-901 is not available yet. Microsoft usually publishes one within about eight weeks of an exam reaching general availability, so keep checking the exam page, but do not wait for it to start studying [1].

Frequently asked questions

How long does it take to study for AI-901?

Most candidates can prepare in about four weeks at roughly one to one and a half hours a day, which matches the volume of Microsoft's own instructor-led course of around 24 hours [7]. If you are new to Python or the Azure portal, add time to the hands-on Foundry weeks. Prior coding experience shortens it considerably.

Can you pass AI-901 in 30 days?

Yes, 30 days is a realistic target if you follow a structured plan and do the hands-on exercises rather than only reading. The exam's larger domain is practical, so the deciding factor is keyboard time in Microsoft Foundry, not calendar length [2]. Compress the schedule if you already code, stretch it if you do not.

Is AI-901 harder than AI-900?

For most people, yes. AI-900 was entirely conceptual, asking you to describe Azure AI services [4]. AI-901 devotes 55-60% of the exam to building apps and agents in Microsoft Foundry and assumes basic Python knowledge [1][2]. It is more demanding, but it also proves a more useful, practical skill set.

What is the passing score for AI-901?

You need 700 on a scale of 1 to 1,000 [3]. That is a scaled score set by Microsoft, not a simple percentage, so 700 is not the same as answering 70% of questions correctly. Partial credit is available on multi-part questions, and wrong answers carry no penalty [3].

Do I need to know Python to pass AI-901?

Some Python literacy is expected. The official audience profile lists "knowledge of Python coding syntax and programming techniques" and familiarity with REST APIs, SDKs, and CLIs [1]. You are not writing complex programs, but you should be able to read and follow the short client applications built in the Foundry learning path modules [9].

Does AI-901 still cover LUIS or Anomaly Detector?

No. Those services do not appear in the AI-901 skills outline, even though the study guide's documentation table still links to them from the old AI-900 template [2]. LUIS retired on 31 March 2026 and Anomaly Detector retires on 1 October 2026, so ignore both and focus on Foundry Tools [13].

Start your AI-901 preparation

A plan only works if you pair it with realistic practice. Rehearse the two-domain outline under timed conditions, then close the gaps the mock results reveal.

Examinotion's AI-901 practice tests mirror the current skills outline and the exam's real pacing, and our AI-901 study guide breaks each domain down further. If you are still choosing which Microsoft AI certification to sit, compare the full catalogue on our Microsoft exams hub. Begin your AI-901 preparation today and give yourself the hands-on confidence this exam now demands.

Sources

1. Microsoft Learn, Exam AI-901: Microsoft Azure AI Fundamentals. https://learn.microsoft.com/en-us/credentials/certifications/exams/ai-901/ (retrieved 8 July 2026)
2. Microsoft Learn, Study guide for Exam AI-901. https://learn.microsoft.com/en-us/credentials/certifications/resources/study-guides/ai-901 (retrieved 8 July 2026)
3. Microsoft Learn, Certification exam scoring and score reports. https://learn.microsoft.com/en-us/credentials/certifications/exam-scoring-reports (retrieved 8 July 2026)
4. Microsoft Learn, Exam AI-900: Microsoft Azure AI Fundamentals (retirement notice). https://learn.microsoft.com/en-us/credentials/certifications/exams/ai-900/ (retrieved 8 July 2026)
5. Microsoft Learn, What is Microsoft Foundry. https://learn.microsoft.com/en-us/azure/foundry/what-is-foundry (retrieved 8 July 2026)
6. Microsoft Learn, What is Azure Content Understanding in Foundry Tools. https://learn.microsoft.com/en-us/azure/ai-services/content-understanding/overview (retrieved 8 July 2026)
7. Microsoft Learn, Course AI-901T00-A: Introduction to AI in Azure. https://learn.microsoft.com/en-us/training/courses/ai-901t00 (retrieved 8 July 2026)
8. Microsoft Learn, AI concepts for developers and technology professionals (learning path). https://learn.microsoft.com/en-us/training/paths/ai-concepts/ (retrieved 8 July 2026)
9. Microsoft Learn, Get started with AI applications and agents on Azure (learning path). https://learn.microsoft.com/en-us/training/paths/get-started-ai-apps-agents/ (retrieved 8 July 2026)
10. Microsoft Learn, Exam duration and exam experience. https://learn.microsoft.com/en-us/credentials/support/exam-duration-exam-experience (retrieved 8 July 2026)
11. Microsoft Learn, Take a Microsoft certification exam online. https://learn.microsoft.com/en-us/credentials/certifications/online-exams (retrieved 8 July 2026)
12. Microsoft Learn, Certifications frequently asked questions. https://learn.microsoft.com/en-us/credentials/certifications/frequently-asked-questions (retrieved 8 July 2026)
13. Microsoft Learn, Language Understanding (LUIS) FAQ (retirement). https://learn.microsoft.com/en-us/azure/ai-services/luis/faq (retrieved 8 July 2026)