Updated AI-901 practice tests are available with 125 questions and answers. Practice with the AI-901 PDF questions and testing engine software, you can prepare for your Microsoft Azure AI Fundamentals exam well.<\/p>\n\n\n\n
Since AI-900 exam, towarding the Microsoft Certified: Azure AI Fundamentals certification, retired on June 30, 2026, you need to complete the AI-901 exam. Compared with AI-900 exam, what changed in the Microsoft Azure AI Fundamentals (Updated Version) AI-901 exam?<\/p>\n\n\n\n The clearest difference is the language used in the objectives. AI-900<\/mark> asked candidates to “describe” and “identify.” AI-901 repeatedly uses “implement,” “create,” “deploy,” “build,” and “interact.” More than half of AI-901 now focuses on practical work in Microsoft Foundry, including prompts, model deployment, agents, SDK-based applications, multimodal models, and Content Understanding.<\/p>\n\n\n\n The most updated AI-901 practice tests from DumpsBase will help you prepare more efficiently. These practice test questions are designed to give you additional opportunities to review exam-related knowledge and work through question-based revision. They can be used after knowing the differences between the AI-900 and AI-901, and completing the learning of the AI-901 exam objectives.<\/p>\n\n\n\n With DumpsBase AI-901 practice tests, you can gain:<\/p>\n\n\n\n You have a Microsoft Foundry project that contains an agent named Agent1. You are developing an application that extracts fields from PDFs by using Azure Content Understanding in Foundry Tools. You need to create an AI agent in Microsoft Foundry that follows a specific role and behavior when responding to users. Your company processes customer support emails. You are developing an application that uses the Azure Language SDK and a Microsoft Foundry resource. Explanation:<\/strong> Comparison<\/th> AI-900<\/mark><\/th> AI-901<\/th><\/tr><\/thead> Current status<\/td> Retired on June 30, 2026<\/td> Active exam with no announced retirement date<\/td><\/tr> Main emphasis<\/td> Understanding and describing AI concepts<\/td> Implementing AI solutions with Microsoft Foundry<\/td><\/tr> Candidate background<\/td> Suitable for technical and nontechnical candidates; data science and software development experience was not required<\/td> Expects basic Python knowledge, familiarity with Azure resources, and foundational technical skills<\/td><\/tr> Exam structure<\/td> Five separate domains covering AI workloads, machine learning, computer vision, NLP, and generative AI<\/td> Two domains: AI concepts and capabilities at 40 to 45%, and Foundry implementation at 55 to 60%<\/td><\/tr> Generative AI<\/td> Identify generative AI concepts and Azure services<\/td> Select and deploy models, write prompts, use the Foundry SDK, and build lightweight applications<\/td><\/tr> Agentic AI<\/td> Not listed as a distinct objective<\/td> Explicitly covers agentic AI and building single-agent solutions<\/td><\/tr> Multimodal AI<\/td> Mainly identifies vision, speech, and NLP capabilities<\/td> Uses multimodal models for text, speech, images, and visual prompts<\/td><\/tr> Information extraction<\/td> Includes document processing as a conceptual workload<\/td> Uses Content Understanding to extract information from documents, images, audio, and video<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Prepare More Efficiently with Updated AI-901 Practice Tests<\/h2>\n\n\n\n
\n
Try 5 Free Demo Questions<\/h2>\n\n\n\n
Question 1<\/h3>\n\n\n\n
You need to ensure that Agent1 always calls an Azure function when the agent responds to user input.
To what should you set tool_choice<\/code> for Agent1?
A. auto<\/code>
B. none<\/code>
C. required<\/code>
Answer:<\/strong> C
Explanation:<\/strong> Microsoft Foundry Agent Service uses tool_choice<\/code> to control tool calling. auto<\/code> allows the model to decide whether to call a tool, while none<\/code> prevents tool calls. Setting it to required<\/code> forces the agent to call one or more tools.<\/p>\n\n\n\nQuestion 2<\/h3>\n\n\n\n
You need to use the Python SDK to submit a PDF for analysis and retrieve the extraction results.
What should you do?
A. Call begin_analyze()<\/code>, and then call poller.result()<\/code> to retrieve the results.
B. Submit the PDF to an analyzer and read the results from the request headers.
C. Use optical character recognition (OCR) to extract text from the PDF and map the fields in the code.
D. Call analyze()<\/code> to return the extracted fields synchronously in the same request.
Answer:<\/strong> A
Explanation:<\/strong> Azure Content Understanding analysis operations are long-running operations in the Python SDK. begin_analyze()<\/code> submits the content and returns a poller. Calling poller.result()<\/code> waits for the operation and retrieves the results. The results are not returned in request headers or synchronously through analyze()<\/code>.<\/p>\n\n\n\nQuestion 3<\/h3>\n\n\n\n
What should you configure?
A. tokens per minute (TPM)
B. system instructions
C. temperature
D. max completion tokens
Answer:<\/strong> B
Explanation:<\/strong> System instructions define an agent’s goals, constraints, role, and behavior. TPM controls throughput, temperature affects response variability, and max completion tokens limits response length.<\/p>\n\n\n\nQuestion 4<\/h3>\n\n\n\n
You need to implement an AI solution that automatically identifies mentions of people, organizations, and locations in the emails.
Which text analysis technique should you use?
A. sentiment analysis
B. Named Entity Recognition (NER)
C. summarization
D. key phrase extraction
Answer:<\/strong> B
Explanation:<\/strong> Named Entity Recognition identifies and categorizes entities in unstructured text, including people, organizations, and locations. Sentiment analysis detects opinion, summarization shortens text, and key phrase extraction identifies important phrases without classifying them as those entity types.<\/p>\n\n\n\nQuestion 5<\/h3>\n\n\n\n
You need to return a list of important topics from unstructured text.
Which Python method should you use?
A. client.recognize_linked_entities()<\/code>
B. client.recognize_entities()<\/code>
C. client.begin_extract_summary()<\/code>
D. client.extract_key_phrases()<\/code>
Answer:<\/strong> D<\/p>\n\n\n\nclient.extract_key_phrases()<\/code> returns key phrases or main topics from unstructured text. The linked-entity method connects detected entities to a knowledge base, the entity-recognition method identifies named entities, and the summary method starts extractive summarization.<\/p>\n\n\n\n