What is the difference between AI-900 and AI-901?

Microsoft retired AI-900 on June 30, 2026, and AI-901 took over. It was not a rename. AI-900 had five conceptual domains and mostly asked you to describe AI workloads. AI-901 has two domains, and 55 to 60 percent of the exam asks you to implement solutions with Microsoft Foundry, including deploying a model and building a small client application. Old AI-900 study notes will not match the current blueprint.

Is the AZ-900 exam the same thing as the Azure AI fundamentals exam?

No, and this trips up a lot of people. The AZ-900 is Microsoft Certified: Azure Fundamentals, a beginner credential about cloud concepts, Azure architecture, and Azure management and governance. It is still live and still listed at 45 minutes. The AI fundamentals credential is a separate exam with a separate code, AI-901. Search results mix the two because both pages use the word fundamentals.

How much does AI-901 cost and does the certification expire?

Microsoft official exam pricing data lists AI-901 at $99 for exams proctored in the United States, with tax not included in the published price. Microsoft states that fundamentals certifications do not expire, so unlike the role-based Azure Administrator Associate credential with its 12 month renewal cycle, this one stays on your transcript until you retire it yourself. There is no annual renewal assessment to pass.

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

You need to read it, not write it. Microsoft says candidates need knowledge of Python syntax and programming techniques, plus familiarity with REST APIs, SDKs, and command line tools. The exam asks you to write prompts, deploy models, and build small client applications in the Foundry portal, and some items use code review. Full time software development experience is not required, but you should be comfortable reading a short script.

Is AI-901 enough for an Azure AI job or do I need AI-200?

AI-901 proves you understand the ideas and can build in the portal, which is enough for many entry level and support roles. AI-200, the Azure AI Engineer Associate credential, is the next step for people who build and ship production AI applications. Most employers treat the fundamentals exam as a screening signal, so plan to take it first and add AI-200 once you have project work to point at.

Microsoft Azure AI Fundamentals Certification Guide

The Microsoft Azure AI Fundamentals certification changed exam code in 2026, and most study guides still point at the old one. The exam you sit now is AI-901, because Microsoft retired AI-900 on June 30, 2026. Microsoft official exam pricing data puts United States fees in two bands, $99 for fundamentals and $165 for associate and expert exams. By the end of this guide you will know which code to study. You will see what the two skill domains cover and how this credential stacks with the Azure AI engineer path.

What happened to the old AI-900 exam?

Microsoft retired AI-900 on June 30, 2026, and its warning box on the exam page says so in plain words. AI-901 replaced it, and the two exams are not the same test with a new label. The old exam used five conceptual domains and asked you to describe AI workloads. The current exam uses two domains, and most of it asks you to build things in Microsoft Foundry.

Two other exam codes get mixed up with this one. The AZ-900 is Microsoft Certified: Azure Fundamentals, a separate beginner credential about cloud concepts and Azure architecture that is still live. The AZ-204 was the Azure Developer Associate exam, and Microsoft has retired that certification along with its renewal assessment. If your study notes say AZ-204 for AI fundamentals, they are describing a code that never covered this material.

If you booked AI-900 before the retirement, check your Microsoft Learn transcript before you sit. Microsoft retired that exam on June 30, 2026, so check that your booking points at AI-901 and not the old code.

What does the Microsoft Azure AI Fundamentals certification test?

Microsoft updated the English exam on April 15, 2026. The study guide lists two domains with weights, and the weights tell you where to spend your time.

Skill domain	Weight
Identify AI concepts and capabilities	40 to 45 percent
Implement AI solutions by using Microsoft Foundry	55 to 60 percent

Identify AI concepts and capabilities

Three blocks sit inside this domain. Responsible AI covers fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. Model components and configurations covers how generative AI models work. It also covers picking a model for a job and choosing deployment options and configuration parameters. AI workloads covers generative and agentic AI, plus text analysis such as keyword extraction, entity detection, sentiment analysis and summarization. Add speech recognition and speech synthesis, computer vision and image generation, and pulling information out of text, images, audio, and video.

Implement AI solutions by using Microsoft Foundry

Microsoft Foundry, the platform Microsoft renamed from Azure AI Studio, is where most of the marks live. The study guide asks you to write effective system and user prompts. Then deploy a model in the Foundry portal, interact with it, and build a lightweight chat client application with the Foundry SDK. It also covers creating and testing a single agent solution, then building a client application for that agent. Add text analysis to a small app, and answer spoken prompts with a deployed multimodal model.

Who should take AI-901 and what do you need to know?

Microsoft describes the audience as people at the beginning of their career in AI solution development. You need conceptual knowledge of AI solutions in Azure plus the basic technical skills to work with them. The exam page also asks for knowledge of Python syntax and programming techniques. Add familiarity with Azure resources, REST APIs, SDKs, and command line tools.

Microsoft lists no prerequisite exam for this credential, and the badge carries the job role of AI Engineer. That makes it the natural first Microsoft exam for a developer, analyst, or product manager. These are people who will work with AI features but not train models from scratch. Data science and software engineering experience is not required, though the exam is not a pure theory paper either.

How is AI-901 scored and how long does it take?

Microsoft scores the exam on a scaled scale out of 1,000 points and you need 700 to pass. That is the same pass mark used across the Microsoft role-based catalog, so it tells you nothing extra about difficulty. The exam page lists fourteen languages and shows no retirement date, so AI-901 is the stable version to plan around. Microsoft also allows a retake 24 hours after a failed attempt, with a longer wait before later attempts.

The exam page lists an instructor-led course called Introduction to AI in Azure. The certification page is thinner. It states that no learning paths or modules are available for this certification yet. The practice assessment sits on AI Skills Navigator, and you must sign in to launch it. The exam sandbox is also free and shows the exact interface you will meet.

What does AI-901 cost in the United States?

Microsoft official exam pricing data lists AI-901 at $99 for exams proctored in the United States. Fundamentals exams sit at that price, while associate and expert exams such as AZ-104, AZ-305, AI-200, and AI-300 sit at $165. Tax is not included, and the price changes by country, so check the live figure when you book.

The Microsoft Azure AI Fundamentals certification never expires. Microsoft states on its renewal page that fundamentals certifications do not expire, unlike associate, expert, and specialty credentials with their 12 month renewal cycle. For a career starter that is the cheapest Microsoft credential you can hold, because you pay once and keep it. Our Azure certification cost guide breaks down every fee line, and the full path breakdown from AZ-900 to AZ-305 prices the wider route.

Practice costs a little, because a free Azure account includes credit but not unlimited usage. The Azure pricing page shows what model deployments and workspaces cost per month. Keeping spend low matters here, since most of the exam can be prepared with a small budget of trial resources.

How does this stack with the Azure AI engineer credentials?

The Microsoft Azure AI Fundamentals certification is the entry point for the AI path in Azure. The next step is AI-200, the Azure AI Engineer Associate credential, which covers building and shipping production AI applications. After that comes AI-300, the Azure AI Solutions Architect Expert credential, where the focus moves to design. Our Microsoft AI certifications guide compares AI-200 and AI-300 in detail.

The infrastructure branch runs in parallel. If your work is Azure platforms rather than models, the Azure Administrator Associate guide covers that path. It also covers the 12 month renewal that this fundamentals credential does not need. Choosing between the two branches is easier once you know which problems you will actually be asked to solve at work.

How should you prepare for AI-901?

Because 55 to 60 percent of the exam is implementation, hands-on time in Microsoft Foundry is worth more than reading. Build one small solution end to end. Deploy a model in the portal, then write a system and a user prompt that produce a reliable answer. After that, build a lightweight chat client with the Foundry SDK. Then create a single agent and test it with a handful of realistic prompts.

Use the free AI Skills Navigator practice assessment to find your weak spot. Use the exam sandbox once to learn where the timer and the navigation buttons are. The official study guide is the blueprint for what to review. If you want a guided route, the Introduction to AI in Azure course covers the conceptual half. Finally, check the best cloud certifications for career growth to see where this exam sits in a wider plan.

Photo: sweetlouise, Pixabay Content License, via Pixabay (https://pixabay.com/photos/computer-school-earphones-books-5130405/)