WebLizard LabsCloud + AI certification training
Menu

AI-901 / 180 original questions

AI-901 Practice Exams: 180 Azure AI Questions

This free AI-901 series contains four 45-question Azure AI Fundamentals practice exams with answer pauses and explanations. The separate 12-question focused lessons are not included in the 180-question full-exam total.

Watch first videoOpen YouTube playlistOpen interactive practice
AI-901 practice exam video thumbnail
4 videos / Beginner

YouTube video playlist

Watch the current AI-901 playlist.

Use the playlist for continuous viewing, or choose a practice exam below to review its length, question range, and key moments before you start. The playlist may also include certification lessons outside this practice-exam set.

Practice exam 1 · 45 questions

AI-901 Practice Exam: Microsoft Azure AI Fundamentals

1h 19m
Watch on YouTube
Key moments and topics
  1. 00:00Cold open / title
  2. 00:17Audience / readiness
  3. 00:40How to use this exam
  4. 01:02Question type primer
  5. 01:22Start exam transition
  6. 01:28Q1: Responsible AI fairness
  7. 02:43Q2: Reliability and safety
  8. 04:30Q3: Privacy, security, and accountability
  9. 06:19Q4: Responsible AI in high-impact workflows
  10. 08:34Q5: Transparency and responsible AI
  11. 10:10Q6: Multimodal model selection
  12. 11:25Q7: Model deployment and SDK configuration
  13. 12:49Q8: Model configuration concepts
  14. 14:38Q9: System instructions and prompt roles
  15. 16:08Q10: Choosing models by modality
  16. 17:49Q11: Recognizing generative AI workloads
  17. 18:54Q12: Text analysis capabilities
  18. 20:18Q13: Speech recognition vs speech synthesis
  19. 22:01Q14: AI workload categories
  20. 23:49Q15: Object detection in computer vision
  21. 24:57Q16: Workload and model selection process
  22. 27:02Q17: Multi-workload AI solution decomposition
  23. 29:04Q18: Image generation workload recognition
  24. 30:07Q19: Agentic AI concepts
  25. 31:55Q20: Deploying and testing models in Foundry
  26. 33:46Q21: Foundry generative AI app workflow
  27. 35:37Q22: Prompt design for reliable support assistants
  28. 37:21Q23: Foundry agents vs simple generative AI interactions
  29. 39:04Midpoint checkpoint
  30. 39:24Q24: Foundry SDK and client app integration
  31. 41:06Q25: Foundry agents with tools and actions
  32. 43:27Q26: Foundry components and workflow roles
  33. 45:23Q27: Grounding with trusted context or retrieval
  34. 47:21Q28: Foundry playground prompt testing
  35. 48:53Q29: Sentiment analysis for customer comments
  36. 50:08Q30: Speech recognition and synthesis
  37. 51:25Q31: Speech, synthesis, and text summarization
  38. 53:10Q32: Text summarization for support notes
  39. 54:21Q33: Voice-enabled support app workflow
  40. 56:30Q34: Multimodal visual input
  41. 57:49Q35: Vision understanding and image generation
  42. 59:16Q36: Matching visual workload capabilities
  43. 01:01:10Q37: Image generation from text prompts
  44. 01:02:17Q38: Responsible image-generation workflow
  45. 01:04:18Q39: Structured document and form extraction
  46. 01:05:56Q40: Content Understanding input types
  47. 01:07:46Q41: Reliable extraction workflow setup
  48. 01:09:42Q42: Structured invoice field extraction
  49. 01:11:31Q43: Multimodal audio and video extraction
  50. 01:13:46Q44: Foundry Tools for repeatable structured extraction
  51. 01:15:43Q45: End-to-end extraction app with human review
  52. 01:18:13Confidence Builder
  53. 01:18:57Thank You / next step

Practice exam 2 · 45 questions

AI-901 Practice Exam 2: Microsoft Azure AI Fundamentals

1h 14m
Watch on YouTube
Key moments and topics
  1. 00:00Welcome and exam strategy
  2. 00:48Q1: Accessible speech assistant
  3. 02:37Q2: Loan-support recommendation monitoring
  4. 04:05Q3: Grounded maintenance assistant
  5. 05:48Q4: Customer-call summarization data boundary
  6. 07:24Q5: AI-generated case summary disclosure
  7. 08:53Q6: Agent tool-action approval
  8. 09:57Q7: Token-by-token generation and context
  9. 11:27Q8: Context limits and unsupported claims
  10. 13:02Q9: Small text model versus multimodal model
  11. 14:36Q10: Temperature and response variability
  12. 16:41Q11: Deployment name, endpoint, and authentication
  13. 18:27Q12: Agentic versus single-turn generative AI
  14. 19:50Q13: Key phrases versus named entities
  15. 21:31Q14: Summarization versus information extraction
  16. 22:52Q15: Live captions and spoken playback
  17. 24:22Q16: Classification, detection, OCR, and generation
  18. 25:55Q17: Text/image/audio/video extraction
  19. 27:34Q18: Generation versus extraction
  20. 28:59Q19: Route multilingual media intake
  21. 30:40Q20: Policy-constrained assistant prompts
  22. 31:59Q21: Grounded answer contract
  23. 33:39Q22: Portal deployment diagnosis
  24. 34:50Q23: Chat client configuration
  25. 36:49Q24: Request/response flow
  26. 38:38Q25: Create and test one tool-using agent
  27. 40:45Q26: Agent tool safety boundary
  28. 42:11Q27: Agent client versus model client
  29. 43:47Q28: Agent conversation and tool result flow
  30. 45:57Q29: Evaluate grounded assistant
  31. 47:30Q30: Authentication failure diagnosis
  32. 49:05Q31: Entity and key-phrase analysis pipeline
  33. 51:08Q32: Batch feedback analysis
  34. 53:03Q33: Spoken question to deployed multimodal response
  35. 54:34Q34: Speech recognition app flow
  36. 56:42Q35: Speech synthesis app flow
  37. 58:32Q36: Image-plus-text inspection client
  38. 60:25Q37: Visual prompt reliability
  39. 61:53Q38: Image-generation request
  40. 63:27Q39: Vision client request/response
  41. 65:32Q40: OCR versus multimodal interpretation route
  42. 67:29Q41: Schema-based onboarding extraction
  43. 69:30Q42: Extract fields from photographed labels
  44. 71:03Q43: Recorded inspection extraction
  45. 72:16Q44: Audio extraction versus transcription
  46. 74:06Q45: End-to-end extraction client

Practice exam 3 · 45 questions

AI-901 Practice Exam 3: Microsoft Azure AI Fundamentals

1h 8m
Watch on YouTube
Key moments and topics
  1. 00:00Introduction and how to use this practice exam
  2. 00:42Q1: Fairness in AI systems
  3. 01:53Q2: Handling uncertain AI recommendations
  4. 03:05Q3: Minimizing retained personal data
  5. 04:18Q4: Accessible alternatives to speech output
  6. 05:23Q5: Disclosing AI use and limitations
  7. 06:47Q6: Accountability for AI services
  8. 08:02Q7: How generative AI produces responses
  9. 09:10Q8: Choosing models by input and output modality
  10. 10:29Q9: Deployment geography and billing requirements
  11. 11:35Q10: Generative AI for original text
  12. 12:40Q11: Text analysis: entities and key concepts
  13. 14:27Q12: Transcribing recorded voice messages
  14. 15:31Q13: Generating images from written briefs
  15. 16:37Q14: Content Understanding field requirements
  16. 18:16Q15: Tokens versus words
  17. 19:43Q16: Model capabilities for application functions
  18. 20:47Q17: Selecting the intended model deployment
  19. 22:11Q18: Analyzing audio and visual content in video
  20. 23:49Q19: Analyzing opinions versus summarizing text
  21. 24:47Q20: Separating system instructions and user requests
  22. 26:25Q21: Testing prompt changes in Foundry
  23. 28:04Q22: Connecting with AIProjectClient in Python
  24. 29:41Q23: Writing prompt-agent instructions
  25. 31:50Q24: Selecting a prompt agent from Python
  26. 33:30Q25: Passing file contents to sentiment analysis
  27. 35:06Q26: Audio input through Chat Completions
  28. 37:32Q27: Speech SDK microphone recognition
  29. 39:54Q28: Building a vision-enabled chat request
  30. 41:18Q29: Text-to-image generation requests
  31. 42:38Q30: Image descriptions and text recognition
  32. 44:09Q31: Reading structured invoice analysis results
  33. 45:20Q32: Defining Content Understanding outputs
  34. 46:42Q33: Analyzing call topics and sentiment
  35. 47:56Q34: Handling asynchronous analysis results
  36. 49:19Q35: Specifying response format in prompts
  37. 50:40Q36: Locating a Foundry project endpoint
  38. 51:46Q37: Reading text from a Responses API result
  39. 53:05Q38: Testing prompt-agent behavior
  40. 55:08Q39: Maintaining context across agent requests
  41. 56:36Q40: Sentiment labels and confidence scores
  42. 58:38Q41: Requesting text and audio responses
  43. 01:00:44Q42: Configuring Speech SDK resource connections
  44. 01:02:22Q43: Follow-up questions about an image
  45. 01:03:50Q44: Extracting fields from varying document layouts
  46. 01:06:26Q45: Submitting a document URL for analysis
  47. 01:08:27Reviewing missed questions and next steps

Practice exam 4 · 45 questions

AI-901 Practice Exam 4: Microsoft Azure AI Fundamentals

1h 11m
Watch on YouTube
Key moments and topics
  1. 00:00How to use this practice exam
  2. 00:42Q01: Evaluating AI outcomes
  3. 02:14Q02: Handling document data
  4. 03:46Q03: Training data and access
  5. 05:19Q04: AI system design
  6. 06:38Q05: Model transparency
  7. 08:09Q06: Model safety
  8. 09:41Q07: Generation variability
  9. 11:11Q08: Multimodal model inputs
  10. 13:33Q09: Capacity and token requirements
  11. 15:07Q10: AI capability selection
  12. 16:10Q11: Matching text-analysis requirements
  13. 18:04Q12: Speech and translation workflows
  14. 19:12Q13: Image and text workflows
  15. 20:37Q14: Content Understanding outputs
  16. 22:08Q15: Prompting and fine-tuning
  17. 23:20Q16: Selecting model capabilities
  18. 24:38Q17: Foundry deployment routes
  19. 25:54Q18: Image extraction and translation
  20. 27:25Q19: Sentiment analysis
  21. 28:31Q20: Prompt design
  22. 30:21Q21: Using deployed models
  23. 32:01Q22: Foundry project endpoints
  24. 33:31Q23: Agents and search tools
  25. 35:23Q24: Agent tool behavior
  26. 37:15Q25: Python text-analysis results
  27. 38:48Q26: Audio model API requests
  28. 40:53Q27: Speech SDK operations
  29. 42:44Q28: Vision model inputs
  30. 44:28Q29: Image generation operations
  31. 45:39Q30: Image client connections
  32. 47:18Q31: Content Understanding schemas
  33. 48:43Q32: Analyzing document images
  34. 50:17Q33: Audio and training requirements
  35. 52:01Q34: HTTP analysis workflows
  36. 53:19Q35: Model prompt configuration
  37. 54:53Q36: Foundry agent prompts
  38. 56:14Q37: Project access and authorization
  39. 57:33Q38: Agent search workflows
  40. 58:57Q39: Agent instructions
  41. 59:59Q40: Text-analysis requirements
  42. 01:02:02Q41: Speech and audio outputs
  43. 01:03:23Q42: Speech result handling
  44. 01:04:59Q43: Image detail and token usage
  45. 01:06:37Q44: OCR output requirements
  46. 01:08:28Q45: Content Understanding API fields
  47. 01:10:54Review and next steps

Cert Lab available

Practice AI-901 online now.

Build focused sessions from the complete AI-901 practice bank, save progress, review explanations, and revisit missed or bookmarked questions.

Open AI-901 Practice

Skills covered in this practice exam.

  • Responsible AI principles, safety, evaluation, transparency, and human review
  • Microsoft Foundry model selection, deployment, prompt design, and application integration
  • Generative AI applications, grounding, tools, and agentic patterns
  • Language, speech, vision, and multimodal workload recognition
  • Information extraction and structured content processing

Use it as a diagnostic.

  1. Read the scenario and commit to an answer before the reveal.
  2. Explain why the distractors fail, not only why the correct answer works.
  3. Record the objective or product area behind every missed question.
  4. Review the current Microsoft Learn study guide, then retry weak sections.

Verify against current public guidance.

These questions are original educational content based on public objectives and official technical documentation. They are not copied assessment items.

Video inventory reviewed October 2, 2026. Exam objectives and product behavior can change; use the current Microsoft Learn study guide as the source of truth.

Official AI-901 exam and study-guide page
WebLizard Labs lizard instructor validating an original practice scenario against current sources

Questions about this AI-901 practice exam.

Are these official AI-901 exam questions?

No. These are original learning questions based on public objectives and Microsoft Learn documentation. They are not copied or leaked assessment items.

How many questions are in this AI-901 practice exam?

The current public WebLizard AI-901 video series contains 180 narrated practice questions with answer reveals, explanations, and exam clues.

How do I check the current AI-901 objectives?

Use the official Microsoft Learn study-guide link on this page. Exam objectives and product behavior can change, so verify the version that applies to your exam date.

How should I use the timed answer windows?

Choose an answer before the reveal, say why the other choices are weaker, and note the objective whenever your reasoning differs from the explanation.