Study guide · Exam overview
The format, passing score, and five content domains of the AWS Certified AI Practitioner exam, using exam guide version 1.1, plus the ideas worth learning before you book AIF-C01.
Exam version notice: This overview follows the AWS Certified AI Practitioner Exam Guide (AIF-C01), version 1.1, published April 30, 2026. Version 1.0 was published March 26, 2026. The exam code did not change. Version 1.1 adds agentic AI, the Model Context Protocol, context engineering, token-based pricing, prompt management, business-alignment metrics, and hallucination grounding. Pricing, languages, and delivery options can change. Confirm them on the AWS certification site before you register. Items tagged Extra are useful context that the exam guide places out of scope.
AWS Certified AI Practitioner is the foundational AI certification. It checks whether you can describe AI, machine learning, and generative AI, pick a fitting use case, and name the AWS service or practice that belongs with that use case. It focuses on practical business applications. It is not a model-building exam.
AWS describes the target candidate as someone with up to 6 months of exposure to AI and ML on AWS who uses these technologies and does not necessarily build them. The exam guide lists work that candidate is not expected to do: coding models or algorithms, feature engineering, hyperparameter tuning, building pipelines, mathematical analysis, implementing security protocols, and writing governance frameworks. If a question asks you to choose a control, it is asking you to recognize the control, not to configure it.
Figures below come from the AIF-C01 exam guide version 1.1 and the AWS certification page reviewed in October 2026.
| Specification | AIF-C01, exam guide 1.1 |
|---|---|
| Level | Foundational. There is no required prerequisite certification. |
| Number of questions | 65. Fifty affect your score. Fifteen are unscored and are not identified on the exam. |
| Question types | Multiple choice, multiple response, ordering (3–5 steps), and matching (3–7 pairs). Unanswered questions are incorrect. There is no penalty for guessing. |
| Length of test | 90 minutes. |
| Passing score | 700 on a scaled range of 100–1,000. Scoring is compensatory: you pass the exam as a whole, not each domain. |
| Recommended experience | Up to 6 months of exposure to AI and ML technologies on AWS. Familiarity with core services such as Amazon EC2, Amazon S3, AWS Lambda, Amazon Bedrock, and Amazon SageMaker AI, plus IAM, the shared responsibility model, and pricing. |
| Exam price | Foundational exams are listed at 100 USD. Taxes and local-currency prices can differ. Confirm the current fee before you book. |
| Validity | Three years. Recertify by passing the latest version of this exam, or by earning the AWS Certified Machine Learning Engineer - Associate. |
A scaled score of 700 is not the same thing as 70 percent of the questions. AWS publishes guide updates about a month before they appear on the exam. Version 1.1 was published April 30, 2026. The AWS certification page also said the Italian and German exams retire after October 15, 2026. Confirm the language list when you register.
These percentages are the share of scored content in exam guide version 1.1. The 200 practice questions on this site follow the same weights.
The vocabulary and the decision of whether a model belongs at all.
How generative systems work, where they fail, and which AWS service hosts them.
The largest domain. Choose the model, the prompt, the customization, and the evaluation.
A smaller domain that is easy to miss if you only study service names.
Recognize the control. You are not asked to implement a security program.
Learn the in-scope services from exam guide version 1.1. The same map is in the core notes and flashcards.
| Need | Service to recognize | Remember |
|---|---|---|
| Call a foundation model | Amazon Bedrock | A managed API for foundation models, including Amazon Nova. You do not train the base model yourself. |
| Answer from your documents | Amazon Bedrock Knowledge Bases | Retrieval Augmented Generation. The model weights stay as they are. |
| Filter model input and output | Amazon Bedrock Guardrails | Denied topics, filters, and sensitive-information checks. Not a substitute for IAM. |
| Store and version prompts | Amazon Bedrock Prompt Management | Reuse a prompt without fine-tuning the model. |
| Run an agent | Amazon Bedrock AgentCore | Identity, policy, tools, and orchestration for agents. Strands Agents is the developer-tool framework. |
| Train your own model | Amazon SageMaker AI | The place for a custom model on your labeled data. JumpStart starts you from a pretrained model. |
| Document a model | Amazon SageMaker Model Cards | Intended use, data notes, metrics, and limitations for a reviewer. |
| Speech, text, and images | Amazon Transcribe, Translate, Polly, Comprehend, Lex, Rekognition, Textract | Each service does one job. Do not swap speech-to-text with text-to-speech. |
| Workplace assistant | Amazon Quick | The in-scope assistant for knowledge work and generative business intelligence. The version 1.1 revision also adds Amazon Q to the in-scope list. |
| Agentic development | Kiro | Listed under Developer Tools, next to Strands Agents. |
| Transform an existing workload | AWS Transform | In the machine-learning in-scope list for transformation work, not a general chat model. |
| Store embeddings | Amazon OpenSearch Service, Amazon Aurora, Amazon Neptune, Amazon RDS for PostgreSQL | Vector search for retrieval. Amazon MemoryDB left the in-scope list in version 1.1. |
| Find sensitive data | Amazon Macie | Discovery in Amazon S3. Encryption keys are AWS KMS. |
| Audit and compliance evidence | AWS CloudTrail, AWS Config, AWS Artifact, Amazon Inspector, AWS Trusted Advisor | API history, configuration, compliance reports, vulnerabilities, and best-practice checks. |
| Extra Threat detection | Amazon GuardDuty, AWS Shield, AWS WAF | Listed out of scope for AIF-C01. Do not treat them as tested services on this exam. |
Cloud foundations such as IAM, Amazon S3, and the shared responsibility model are assumed knowledge. They are covered in the AWS Cloud Practitioner notes if those names are still new.
Multiple choice has one correct response and three distractors. Multiple response has two or more correct responses out of five or more options. The stem tells you how many to choose. Select every response that completes the statement, and nothing extra. You receive credit only when the whole set is correct.
Ordering gives three to five steps. You must place every step correctly. Matching pairs three to seven prompts with responses. You must match every pair. A blank answer is incorrect. There is no published penalty for a wrong guess, so answer every item.
Distractors are nearby ideas. Retrieval is not fine-tuning. Precision is not recall. Amazon Polly is not Amazon Transcribe. Read the noun in the question: train, retrieve, prompt, filter, or govern.
A practical sequence if the service names still blur together. Spend the most time on domain 3, then domain 2.
Say how AI, ML, deep learning, generative AI, and agentic AI differ, and which learning type fits labeled data, unlabeled groups, or a reward.
If the business needs one exact result, or a simple rule already works, a prediction is the wrong tool. Then choose a traditional model when you need explainability.
Bedrock for a foundation-model API. Knowledge Bases for retrieval. Guardrails for filters. SageMaker AI for a model you train. One speech or vision service per job.
Prompt and in-context examples first, then retrieval, then fine-tuning, then training from scratch. Distillation is how you make later inference cheaper.
Name the risk, the dataset problem, and the AWS feature that records, filters, encrypts, or audits it. Ground answers instead of raising the temperature.
Use the five flashcard decks, then the practice exams. Review every explanation and return to domains below 75 percent. Aim for a steady 85 percent before you book.
The overview, core notes, flashcards, and practice exams are all published.
Exam format, the 700 passing score, domain weights, and a study plan.
Reference notes for learning types, generative AI, prompts, evaluation, and governance.
50 recall cards in five decks, plus memory notes for models, prompts, and controls.
200 original questions with custom exams, explanations, and a score report by domain.
The current exam code is AIF-C01. These notes follow exam guide version 1.1, published April 30, 2026. There is no newer exam code. Confirm the guide version on the AWS certification page before you book.
Results use a scaled score from 100 to 1,000. The minimum passing score is 700. Scoring is compensatory, so you do not need a passing score in every domain. A scaled score is not the percentage of questions you answered correctly.
The exam has 65 questions and a 90-minute time limit. Fifty questions affect your score. Fifteen are unscored and are not labeled. Question types are multiple choice, multiple response, ordering, and matching.
There is no required prerequisite. AWS describes the target candidate as someone with up to 6 months of exposure to AI and ML technologies on AWS. The candidate uses AI solutions and is not expected to build models, tune hyperparameters, or write production pipelines.
The certification is valid for three years. You can recertify by passing the latest version of this exam, or by earning the AWS Certified Machine Learning Engineer - Associate. Confirm the current policy with AWS before your expiration date.
No. The practice questions on this site are original. They are not AWS exam items, and a practice score does not predict a result.
Service names and prices change. Use the current AWS exam guide and the AWS certification site as the authority when they differ from these notes.
Useful companions while you study.
Generate strong random passwords with custom length, characters, symbols, and security options.
Convert structured data between JSON and YAML with formatting and validation.
Encode text and files to Base64 or decode Base64 data with UTF-8 support.
Compare two text blocks and highlight added, removed, and changed content.
Exam versions, objectives, prices, languages, and retirement dates can change. Verify the details on the AWS certification page before you register.
Certicab provides general informational, educational, and convenience resources. Calculations, conversions, estimates, and learning materials may contain errors or become outdated. Financial, tax, medical, legal, and travel information is not professional advice. Verify important results and current requirements with qualified professionals or authoritative sources. Protect sensitive files and personal information, review each tool’s privacy limitations, and use only content you are authorized to process. You are responsible for how you use and share results. Study resources are independent and do not guarantee exam success or imply certification-provider endorsement. Amazon Web Services, AWS, and related marks are trademarks of Amazon.com, Inc. or its affiliates. Certicab is not affiliated with, endorsed by, or sponsored by Amazon.
Certi