AWS Certified AI Practitioner (AIF-C01)
Amazon Web Services
Complete guide to passing the AWS Certified AI Practitioner (AIF-C01) exam on your first attempt.
$100
Not published
3 years
Global
Amazon Web Services
$102k-$175k
Are you ready for AWS Certified AI Practitioner (AIF-C01)?
Loading quiz...
Complete Overview
AWS Certified AI Practitioner (AIF-C01) is a foundational-level Amazon Web Services certification that tests knowledge of artificial intelligence, machine learning and generative AI concepts together with the AWS services that deliver them. It costs 100 USD, runs 90 minutes, contains 65 questions, and is aimed at people who use AI and ML solutions on AWS without necessarily building them, including business analysts, IT support staff, marketing professionals, product managers and sales professionals. AWS states the target candidate has up to six months of exposure to AI and ML technologies on AWS.
The exam covers five content domains with published weightings: Fundamentals of AI and ML at 20 percent, Fundamentals of Generative AI at 24 percent, Applications of Foundation Models at 28 percent, Guidelines for Responsible AI at 14 percent, and Security, Compliance, and Governance for AI Solutions at 14 percent. Only 50 of the 65 questions affect the score. The other 15 are unscored pretest items that AWS uses to evaluate future questions, and they are not marked on screen. Results are reported on a scaled range of 100 to 1,000 with a minimum passing score of 700, and the scoring model is compensatory, so a weak domain can be offset by a strong one.
Question formats include multiple choice with one correct answer among four options, multiple response with two or more correct answers among five or more options, ordering questions with a list of three to five responses, matching questions pairing responses against three to seven prompts, and case studies where one scenario carries two or more separately scored questions. Unanswered questions are scored as incorrect and there is no guessing penalty.
Candidates sit the exam at a Pearson VUE test centre or online with a remote proctor. AWS offers the exam in Arabic, English, French (France), German, Italian, Japanese, Korean, Portuguese (Brazil), Spanish (Latin America), Spanish (Spain), Simplified Chinese and Traditional Chinese. Non-native English speakers taking the English version can request an ESL +30 accommodation that adds 30 minutes, and that request is made once and then applies to all future AWS exam registrations.
The certification is valid for three years. Holders recertify by passing the current version of AIF-C01 or by earning AWS Certified Machine Learning Engineer - Associate, which recertifies AI Practitioner automatically. A failed attempt requires a 14 calendar day wait before rebooking, there is no cap on the number of attempts, and the full registration fee applies each time. After a pass, the same exam cannot be retaken for two years.
AWS notes that the Cloud Practitioner content outline contains a single task statement related to AI, while the entire AI Practitioner outline is about AI, ML and generative AI. Model building, feature engineering, hyperparameter tuning, pipeline deployment and statistical analysis of models are explicitly out of scope for the target candidate.
Why Get AWS Certified AI Practitioner (AIF-C01) Certified?
AWS prices the exam at 100 USD, the foundational tier, against 150 USD for associate exams and 300 USD for professional and specialty exams.
The exam is 90 minutes for 65 questions, about 83 seconds each. Associate exams carry the same 65 questions over 130 minutes, so this paper runs the tighter clock.
Generative AI content makes up 52 percent of the scored exam once Domain 2 (24 percent) and Domain 3 (28 percent) are combined.
Passing recertifies automatically for three years, and earning AWS Certified Machine Learning Engineer - Associate resets that clock without resitting AIF-C01.
Twelve exam languages are offered, so candidates in Japan, Korea, Brazil, Germany and Mainland China can test in their own language.
Only 50 questions are scored, so 15 unscored items cannot pull a borderline candidate below the 700 cut score.
AWS lists business analysts, marketing professionals and sales professionals as target candidates, which makes this one of the few AI certifications aimed outside engineering.
Exam Format & Structure
Duration
90 minutes
Questions
65
Passing Score
700 on a scaled range of 100 to 1,000
Question Types
- Multiple choice: one correct response and three distractors
- Multiple response: two or more correct responses out of five or more options
- Ordering: place three to five responses in the correct sequence
- Matching: pair responses against three to seven prompts
- Case study: one scenario with two or more separately scored questions
Delivery Method
Pearson VUE test centre or online proctored exam
- Fifteen of the 65 questions are unscored pretest items and are not identified during the exam.
- Unanswered questions are scored as incorrect and there is no penalty for guessing.
- Candidates who register for a localised exam can toggle to view questions in English during the exam.
- An ESL +30 accommodation adds 30 minutes for non-native English speakers sitting the English version.
How scoring works
Score scale
100 to 1,000 scaled score
Score needed to pass
700
Roughly what that means raw
AWS does not publish a raw score to scaled score conversion table for AIF-C01.
When results arrive
Detailed results appear within five business days in the AWS Certification account under Exam History.
How the scale is built
AWS scores 50 of the 65 questions and discards 15 unscored pretest items. The raw result is converted to a 100 to 1,000 scaled score so that forms of slightly different difficulty are comparable. Scoring is compensatory, meaning no domain has its own minimum and a weak domain can be offset by a strong one. The result is reported as pass or fail.
- The score report may include a section-level table of classifications showing relative strengths and weaknesses, and AWS advises caution when interpreting it.
- Unanswered questions are scored as incorrect.
- The exam is scored against a minimum standard set by AWS subject matter experts, not against how other candidates performed on the same form.
- The 50 scored and 15 unscored split does not apply to beta versions of an AWS exam, which the exam guide footnotes separately.
Pacing and time budget
65 questions in 90 minutes gives you About 83 seconds per question per question.
| At this point | You should have answered |
|---|---|
| 20 min | 15 |
| 45 min | 33 |
| 65 min | 48 |
| 82 min | 65 |
- Leaving eight minutes at the end covers a review of flagged items without rushing the final questions.
- Case study questions share one scenario, so read the scenario once and answer its questions consecutively rather than returning to it repeatedly.
- Ordering and matching items take longer than multiple choice, so bank time on definition questions early.
- Candidates granted ESL +30 have 120 minutes, which lifts the per-question budget to about 110 seconds.
- Never leave a question blank at the end, because a blank scores the same as a wrong answer and a guess costs nothing.
Where the marks are
| Topic | Weight | Why it scores |
|---|---|---|
| Applications of Foundation Models | 28% | The largest domain. RAG design, inference parameters, prompt engineering techniques, fine-tuning tradeoffs and evaluation metrics all sit here, and its four task statements cover more objectives than any other domain. |
| Fundamentals of Generative AI | 24% | Second largest, and it supplies the vocabulary that Domain 3 questions assume. Tokens, embeddings, transformer LLMs, diffusion models and the Amazon Bedrock service boundary recur in scenario stems across the exam. |
| Fundamentals of AI and ML | 20% | Definition-heavy and fast to answer, which makes it the best place to bank time. Supervised versus unsupervised learning, regression versus classification, and AWS managed AI services are direct recall. |
| Guidelines for Responsible AI | 14% | Small but predictable. Guardrails for Amazon Bedrock, SageMaker Clarify, Model Cards, Amazon Augmented AI and the overfitting-underfitting distinction cover most of what is asked. |
| Security, Compliance, and Governance for AI Solutions | 14% | Rewards knowing which AWS governance service produces which artifact. AWS Config, Audit Manager, Artifact, CloudTrail, Inspector and Trusted Advisor are separated by output, not by concept. |
| RAG versus fine-tuning cost tradeoff | Within the 28% Domain 3 | Task statement 3.1 asks candidates to explain cost tradeoffs across pre-training, fine-tuning, in-context learning and RAG. Scenarios usually give a budget or a data freshness constraint that decides the answer. |
| AWS vector storage options | Within the 28% Domain 3 | The exam guide names five by service: Amazon OpenSearch Service, Amazon Aurora, Amazon Neptune, Amazon DocumentDB and Amazon RDS for PostgreSQL. Pure recall, and easy marks once memorised. |
| Prompt risk vocabulary | Within the 28% Domain 3 | Exposure, poisoning, hijacking and jailbreaking are named explicitly in task statement 3.2 and reappear in Domain 5 security questions about prompt injection. |
The numbers
100 USD
Exam fee
Source: AWS Certified AI Practitioner exam page, aws.amazon.com/certification
65 questions in 90 minutes
Questions and duration
Source: AWS Certified AI Practitioner exam page, aws.amazon.com/certification
50 scored, 15 unscored
Scored questions
Source: AWS Certified AI Practitioner (AIF-C01) Exam Guide, Version 1.4
700 on a 100 to 1,000 scale
Passing score
Source: AWS Certified AI Practitioner (AIF-C01) Exam Guide, Version 1.4
14 calendar days, no attempt limit
Retake wait
Source: AWS Certification FAQ, aws.amazon.com/certification/faqs
Within five business days
Result timing
Source: AWS Certification FAQ, aws.amazon.com/certification/faqs
If you fail
Wait before retaking
14 calendar days after a failed attempt
Attempt limit
No limit on the number of attempts
Retake fee
Full 100 USD registration fee for every attempt
Once you have passed an exam, you cannot retake the same exam for two years. An appointment can be rescheduled up to 24 hours before the exam time and only twice. Cancelling inside 24 hours, or failing to appear, forfeits the exam fee, and the test delivery provider charges the full delivery fee with no refund.
Exam Domains & Topics
Covers baseline vocabulary and the machine learning lifecycle. Candidates define AI, ML, deep learning, neural networks, computer vision, NLP, bias, fairness, fit and large language models, distinguish supervised, unsupervised and reinforcement learning, describe batch and real-time inferencing, and map AWS managed services to each stage of an ML pipeline from data collection through monitoring and retraining.
Key Topics to Master:
- Labeled versus unlabeled, tabular, time-series, image and text data
- Supervised, unsupervised and reinforcement learning
- Batch versus real-time inferencing
- Regression, classification and clustering technique selection
- Amazon Transcribe, Translate, Comprehend, Lex and Polly capabilities
- SageMaker Data Wrangler, Feature Store and Model Monitor
- MLOps concepts including retraining and production readiness
- Accuracy, AUC and F1 score against business metrics such as ROI
Tests generative AI building blocks and the AWS services that host them. Candidates explain tokens, chunking, embeddings, vectors, transformer-based LLMs, multi-modal models and diffusion models, walk the foundation model lifecycle from data selection through feedback, weigh advantages against hallucination and nondeterminism, and compare cost tradeoffs across Amazon Bedrock, SageMaker JumpStart, Amazon Q and PartyRock.
Key Topics to Master:
- Tokens, chunking, embeddings and vector representations
- Transformer-based LLMs, multi-modal models and diffusion models
- Foundation model lifecycle: pre-training, fine-tuning, evaluation, deployment
- Hallucinations, interpretability limits and nondeterminism
- Amazon Bedrock, SageMaker JumpStart, Amazon Q, PartyRock
- Token-based pricing and provisioned throughput tradeoffs
- Business value metrics: conversion rate, average revenue per user, customer lifetime value
The largest scored block. Candidates select pre-trained models against cost, modality, latency, model size and context length, explain how temperature and output length change responses, define Retrieval Augmented Generation and name AWS vector stores, choose prompt engineering techniques, describe fine-tuning and data preparation including RLHF, and evaluate model output with ROUGE, BLEU and BERTScore.
Key Topics to Master:
- Model selection criteria: cost, modality, latency, context length
- Inference parameters including temperature and output length
- Retrieval Augmented Generation with Amazon Bedrock knowledge bases
- Vector storage in OpenSearch Service, Aurora, Neptune, DocumentDB, RDS for PostgreSQL
- Zero-shot, single-shot, few-shot and chain-of-thought prompting
- Prompt risks: exposure, poisoning, hijacking, jailbreaking
- Instruction tuning, transfer learning and continuous pre-training
- ROUGE, BLEU, BERTScore and human evaluation
- Agents for Amazon Bedrock in multi-step tasks
Covers fairness, transparency and legal exposure. Candidates identify responsible AI features such as inclusivity, robustness, safety and veracity, apply Guardrails for Amazon Bedrock, recognise intellectual property and biased-output risks, describe dataset characteristics that reduce bias, separate overfitting from underfitting, and explain where SageMaker Clarify, Model Cards and Amazon Augmented AI fit.
Key Topics to Master:
- Bias, fairness, inclusivity, robustness, safety and veracity
- Guardrails for Amazon Bedrock
- Legal risks: IP infringement claims, biased outputs, end user risk
- Curated, diverse and balanced datasets
- Bias and variance effects, overfitting and underfitting
- SageMaker Clarify, SageMaker Model Monitor, Amazon Augmented AI
- SageMaker Model Cards and explainability tradeoffs
- Human-centered design for explainable AI
Tests how AI workloads are secured and governed on AWS. Candidates apply IAM roles and policies, encryption, Amazon Macie and AWS PrivateLink, document data lineage and cataloguing, defend against prompt injection, identify ISO and SOC compliance standards and algorithm accountability laws, and use AWS Config, Audit Manager, Artifact, CloudTrail, Inspector and Trusted Advisor for governance evidence.
Key Topics to Master:
- IAM roles, policies and permissions for AI workloads
- Encryption at rest and in transit, AWS PrivateLink, Amazon Macie
- Data lineage, data cataloguing and source citation
- Prompt injection, threat detection and vulnerability management
- ISO and SOC standards, algorithm accountability laws
- AWS Config, Audit Manager, Artifact, CloudTrail, Inspector, Trusted Advisor
- Data residency, retention, logging and lifecycle governance
- Generative AI Security Scoping Matrix
Recommended Study Plan
- 1Read the AIF-C01 exam guide end to end and note the five domain weights
- 2Write flashcards for AI, ML, deep learning, neural network, model, algorithm, inference and LLM
- 3Learn the difference between supervised, unsupervised and reinforcement learning with one example each
- 4Create an AWS account and open the Amazon Bedrock console to see available foundation models
- 5Take the free AWS Certification Official Practice Question Set to calibrate the question style
- 1Map the ML pipeline stages: collection, EDA, pre-processing, feature engineering, training, tuning, evaluation, deployment, monitoring
- 2Match each stage to SageMaker Data Wrangler, Feature Store, Model Monitor and Clarify
- 3Run a short demo in Amazon Comprehend and Amazon Transcribe to see managed AI in action
- 4Learn accuracy, precision, recall, F1 and AUC well enough to say when each is the right metric
- 5List three business cases where a rules engine beats an ML model
- 1Drill regression versus classification versus clustering across 20 short scenarios
- 2Compare batch inference and real-time endpoints on cost and latency
- 3Summarise what Amazon Translate, Lex, Polly, Rekognition and Textract each do in one line
- 4Review MLOps ideas: repeatable pipelines, technical debt, model drift and retraining triggers
- 5Answer 30 Domain 1 practice questions and log every wrong answer with the reason
- 1Define tokens, chunking, embeddings and vectors and explain how a vector search returns results
- 2Trace the foundation model lifecycle from data selection to deployment and feedback
- 3List four generative AI limitations: hallucination, interpretability, inaccuracy and nondeterminism
- 4Build one app in PartyRock to see prompt-driven application assembly
- 5Read the Amazon Bedrock user guide sections on model access and inference
- 1Compare Amazon Bedrock, SageMaker JumpStart and Amazon Q on who each is built for
- 2Learn token-based pricing against provisioned throughput and when each is cheaper
- 3Note which Bedrock features are regional and how that affects data residency answers
- 4Connect generative AI outcomes to business metrics such as conversion rate and ARPU
- 5Answer 30 Domain 2 practice questions and rewrite the ones you missed as flashcards
- 1Rank pre-trained model selection criteria: cost, modality, latency, size, context length, customisation
- 2Explain Retrieval Augmented Generation in three sentences without using the word retrieval twice
- 3Memorise the AWS vector store options: OpenSearch Service, Aurora, Neptune, DocumentDB, RDS for PostgreSQL
- 4Build a Bedrock knowledge base against a small document set
- 5Test how temperature changes output on the same prompt across three values
- 1Write one worked example each of zero-shot, single-shot, few-shot and chain-of-thought prompting
- 2Practise negative prompts and prompt templates on a single business task
- 3Learn the four prompt risks: exposure, poisoning, hijacking, jailbreaking
- 4Apply Guardrails for Amazon Bedrock to block a category and observe the response
- 5Answer 25 prompt engineering questions and time yourself at 80 seconds each
- 1Rank pre-training, fine-tuning, in-context learning and RAG by cost and effort
- 2Describe instruction tuning, domain adaptation and continuous pre-training in one line each
- 3Learn what RLHF adds to fine-tuning data preparation
- 4Memorise ROUGE for summarisation, BLEU for translation and BERTScore for semantic similarity
- 5Explain when human evaluation beats a benchmark dataset
- 1List the responsible AI features: bias, fairness, inclusivity, robustness, safety, veracity
- 2Separate overfitting from underfitting and name the fix for each
- 3Map SageMaker Clarify, Model Monitor and Amazon Augmented AI to bias detection and human review
- 4Read what a SageMaker Model Card records and why it supports transparency
- 5Write down three legal risks of generative AI output that AWS names in the exam guide
- 1Apply the AWS shared responsibility model to a Bedrock workload
- 2Distinguish AWS Config, Audit Manager, Artifact, CloudTrail, Inspector and Trusted Advisor by output
- 3Learn where Amazon Macie and AWS PrivateLink fit in an AI data pipeline
- 4Explain prompt injection as a security risk rather than a prompt quality problem
- 5Read the Generative AI Security Scoping Matrix and place two workloads on it
- 1Sit a full 65-question practice exam under a 90 minute timer
- 2Score by domain and rank the five domains by weakness
- 3Spend two sessions only on the weakest domain
- 4Redo every question missed in weeks 3, 5 and 7
- 5Practise the ordering and matching formats, which most candidates see least often
- 1Sit a second full practice exam and confirm you finish with 10 minutes spare
- 2Review the in-scope AWS services appendix in the exam guide one service at a time
- 3Book the exam slot and confirm your two forms of ID match your AWS Certification account name
- 4If testing online, run the Pearson VUE system test on the same machine and network
- 5Rest the day before rather than adding a new topic
Ready to pass AWS Certified AI Practitioner (AIF-C01)?
Get 500+ practice questions, video walkthroughs, and a pass guarantee.
Best Study Resources
AWS Certified AI Practitioner (AIF-C01) Exam Guide
Official exam guideThe 19-page PDF that publishes the five domain weights, the five question formats, the 50 scored and 15 unscored split, the 100 to 1,000 scale, and the in-scope and out-of-scope AWS service appendix.
Free
AWS Skill Builder Exam Prep Plan: AWS Certified AI Practitioner
Structured learning planAWS four-step prep plan covering exam familiarisation, knowledge refresh, review and readiness assessment, with digital courses and flashcards mapped to each task statement.
Free tier available
AWS Certification Official Practice Question Set
Practice questionsAWS-authored sample questions in the real exam formats, useful early to learn how AWS phrases distractors before you have covered every domain.
Free
AWS Certification Official Practice Exam
Timed practice examFull-length timed practice exam scored by domain, which is the closest available signal on readiness because AWS does not release pass rates or historical score data.
Included with AWS Skill Builder individual subscription at $29 per month
Amazon Bedrock User Guide
Official documentationPrimary reference for foundation model access, inference parameters, knowledge bases, agents and guardrails, all of which sit inside the 28 percent Applications of Foundation Models domain.
Free
PartyRock, an Amazon Bedrock Playground
Hands-on sandboxBrowser-based app builder that lets non-developers assemble generative AI apps from prompts, which makes prompt templates, chaining and model behaviour concrete rather than theoretical.
Free
AWS Well-Architected Machine Learning Lens
WhitepaperAWS design guidance across the ML lifecycle covering operational excellence, security, reliability, performance and cost, which lines up with Domain 1 and Domain 5 task statements.
Free
Generative AI Security Scoping Matrix
FrameworkAWS framework that classifies generative AI workloads into five scopes from consumer app to self-trained model, named directly in Domain 5 task statement 5.2 on governance protocols.
Free
Amazon SageMaker AI Developer Guide
Official documentationReference for JumpStart, Clarify, Model Monitor, Model Cards, Data Wrangler and Feature Store, the SageMaker components named across Domains 1, 3 and 4 of the exam guide.
Free
AWS Certification policies and before-testing guidance
Policy referenceAWS pages covering the 14 day retake wait, five business day result timing, two-form ID requirement, ESL +30 accommodation and Pearson VUE accommodation process.
Free
Common Mistakes to Avoid
Studying model training in depth because the exam mentions fine-tuning.
The exam guide lists developing or coding models, feature engineering, hyperparameter tuning and pipeline building as out of scope. Learn what fine-tuning costs and when to choose it, not how to run a training job.
Treating all five domains as equal and spending equal hours on each.
Applications of Foundation Models is 28 percent and Fundamentals of Generative AI is 24 percent. Together they are 52 percent of the scored content. Responsible AI and Security are 14 percent each.
Confusing Amazon Bedrock, SageMaker JumpStart and Amazon Q when a scenario names a user persona.
Anchor each to its audience: Bedrock for API access to third-party and Amazon foundation models, JumpStart for pre-built models and notebooks inside SageMaker, Amazon Q for a managed assistant experience.
Answering RAG questions by describing fine-tuning.
RAG retrieves external content at inference time and leaves model weights unchanged. Fine-tuning changes weights and needs curated training data. Cost, freshness of data and governance all differ, and exam scenarios turn on that difference.
Skipping the ordering and matching question formats during practice.
Ordering questions need all three to five items in the right sequence and matching questions need every pair correct for any credit. Partial knowledge scores zero, so drill sequences such as the ML pipeline and the foundation model lifecycle until they are automatic.
Assuming a wrong answer in one domain can be recovered by leaving it blank.
Unanswered questions are scored as incorrect and there is no guessing penalty. Mark and return, but never submit the exam with a blank.
Learning evaluation metrics as a single list without task association.
Tie ROUGE to summarisation, BLEU to translation and BERTScore to semantic similarity, and keep accuracy, AUC and F1 with classical ML. Scenario questions name the task first and expect the metric second.
Treating prompt injection as a prompt writing problem.
Prompt injection, poisoning, hijacking and jailbreaking sit in the security and prompt-risk objectives. Pair them with Guardrails for Amazon Bedrock, input validation and least-privilege IAM rather than with prompt style advice.
Booking the exam without checking the name on the AWS Certification account.
AWS requires valid government-issued identification, and a secondary form of ID is also required. Update the account name to match the ID before scheduling, because a mismatch on the day forfeits the appointment.
Preparing with third-party question dumps rather than the AWS exam guide appendix.
The exam guide appendix lists in-scope and out-of-scope AWS services explicitly. Working that list is a faster way to close gaps than memorising questions that may reference retired service names.
Exam Day Tips
- 1
Budget 83 seconds per question, since 90 minutes divided by 65 questions leaves no slack for long deliberation.
- 2
Answer every question before submitting, because unanswered items are scored as incorrect and guessing carries no penalty.
- 3
Flag case study scenarios and return to them together, as the scenario text is shared and rereading it once serves several questions.
- 4
On multiple response items count the required number of answers first, since selecting too few scores zero even if the choices are right.
- 5
Bring two forms of identification to a test centre, one of them government issued, and check the names match your AWS Certification account.
- 6
For online proctoring, clear the desk completely, close every application before the check-in, and test the webcam and microphone on the same network you will use.
- 7
No personal items, notes, calculators or scratch paper are permitted, so build the mental checklists you need during study rather than planning to write them down.
- 8
If the exam is localised, use the English toggle when a translated option reads ambiguously, because the English source wording is the one the item was written in.
- 9
Work the exam guide appendix before the day, because it names the in-scope AWS services by category, from Amazon Bedrock and Amazon Q through AWS Config, CloudTrail and Trusted Advisor.
- 10
If you requested ESL +30, confirm the accommodation is attached to the appointment before check-in rather than raising it with the proctor.
Career Paths & Salary Ranges
AI and cloud solutions consultant
Advises business units on where generative AI fits, scopes Amazon Bedrock and Amazon Q pilots and writes the cost cases. BLS reports a May 2025 median annual wage of $101,860 for management analysts, with the 90th percentile at $171,640.
$102k-$172k
Data scientist
Builds and evaluates models, selects metrics and validates foundation model output against benchmarks. BLS reports a May 2025 median annual wage of $120,230 for data scientists, with the 90th percentile at $199,130.
$120k-$199k
Application developer building on Amazon Bedrock
Writes the application layer around foundation models, including retrieval pipelines, agents and guardrail configuration. BLS reports a May 2025 median annual wage of $135,980 for software developers, with the 90th percentile at $214,670.
$136k-$215k
AI product marketing manager
Positions AI features, runs competitive analysis on model capabilities and translates model limits into customer-facing claims. BLS reports a May 2025 median annual wage of $78,760 for market research analysts and marketing specialists, with the 90th percentile at $155,480.
$79k-$155k
IT manager leading AI adoption
Owns the AI service portfolio, governance sign-off and vendor decisions across a business unit. BLS reports a May 2025 median annual wage of $175,140 for computer and information systems managers, with the 90th percentile at $297,510.
$175k-$298k
Prerequisites & Requirements
- No formal prerequisites. AWS does not require prior certification, training or work experience to register.
- AWS recommends up to six months of exposure to AI and ML technologies on AWS.
- Familiarity with the core AWS services named in the exam guide: Amazon EC2, Amazon S3, AWS Lambda, Amazon Bedrock and Amazon SageMaker AI.
- Familiarity with the AWS shared responsibility model for security and compliance.
- Familiarity with AWS Identity and Access Management for controlling access to resources.
- Familiarity with AWS global infrastructure, including Regions, Availability Zones and edge locations.
- Familiarity with AWS service pricing models.
- Candidates new to AWS are directed to AWS Cloud Practitioner Essentials or AWS Technical Essentials first.
- A valid government-issued ID establishing residence in a non-sanctioned country, plus a secondary form of ID, is required to test.
Frequently Asked Questions
How much does the AWS Certified AI Practitioner exam cost?
The exam costs 100 USD. AWS publishes local pricing separately on its exam pricing page, where the amount charged in other currencies follows AWS foreign exchange rates and local taxes may be added at checkout. Once you hold any AWS Certification, your AWS Certification account carries a 50 percent discount voucher that can be applied to a future exam, valid for half the exam price at the time of purchase.
What is the passing score?
700 out of a scaled range of 100 to 1,000. The score is not a percentage of questions answered correctly. AWS uses scaled scoring so that results are comparable across exam forms that differ slightly in difficulty, and the exam is reported as a pass or fail.
How many questions are scored?
50 of the 65 questions affect your score. The remaining 15 are unscored pretest items that AWS uses to evaluate questions for future use, and they are not identified during the exam.
What happens if I fail?
You must wait 14 calendar days before retaking the exam. There is no limit on the number of attempts, but the full registration fee applies to every attempt. A failed attempt does not affect any other AWS certification you already hold.
When do I get my result?
Detailed results appear within five business days in your AWS Certification account under Exam History, where the score report can be downloaded as a PDF. The report gives the scaled score and a section-level table of classifications rather than a per-question breakdown.
How long is the certification valid?
Three years. You can recertify by passing the current version of the AIF-C01 exam, or by earning AWS Certified Machine Learning Engineer - Associate, which recertifies AWS Certified AI Practitioner automatically.
Can I retake the exam after passing it?
No, not for two years. AWS states that once you have passed an exam you will not be able to retake the same exam for two years. Recertification is handled separately: passing the current version of AIF-C01 when your three years are nearly up, or earning AWS Certified Machine Learning Engineer - Associate, renews the credential.
What identification do I need?
A valid government-issued identification establishing residence in a non-sanctioned country, plus a secondary form of ID. If you do not hold a qualifying government-issued ID from the country you are testing in, an international travel passport from your country of citizenship must be presented as the primary ID, and a secondary ID is still required. ID rules differ between online and in-person testing, so read the requirement page before booking.
Can I take the exam online?
Yes. Online proctoring is available for all AWS Certification exams through Pearson VUE, and most appointments are available 24 hours a day, seven days a week. You take the exam from a private space on your own computer while a proctor monitors screen share and webcam.
What materials am I allowed during the exam?
None. AWS exams are closed book, with no notes, printed material, calculators or scratch paper permitted, and no access to documentation. Online proctored candidates must clear the workspace before the proctor releases the exam.
Are accommodations available?
Yes. Reasonable accommodations for documented disabilities are arranged in advance with the test delivery provider and must be requested before scheduling each exam. Non-native English speakers can request ESL +30, which adds 30 minutes to an exam taken in English, and that request is made once and then applies to all future AWS exam registrations.
Which languages is the exam offered in?
Twelve: Arabic, English, French (France), German, Italian, Japanese, Korean, Portuguese (Brazil), Spanish (Latin America), Spanish (Spain), Simplified Chinese and Traditional Chinese. Candidates who register for a localised exam can toggle to view questions in English during the exam.
How is this different from AWS Certified Cloud Practitioner?
Cloud Practitioner is a foundational overview of the AWS Cloud and its content outline contains only one task statement related to AI. The entire AI Practitioner outline is about AI, ML and generative AI. Both sit at the foundational tier and both cost 100 USD, so candidates sometimes take both.
Should I take AI Practitioner or Machine Learning Engineer - Associate?
Take AI Practitioner if you use AI and ML solutions without building them. AWS explicitly places model coding, feature engineering, hyperparameter tuning and pipeline deployment out of scope for the AI Practitioner target candidate, and those tasks are the substance of the associate-level Machine Learning Engineer exam.
Does AWS publish a pass rate?
No. AWS does not publish pass rates for any of its certification exams, and it does not release score distributions. The only readiness signal from AWS itself is the Official Practice Exam on AWS Skill Builder, which is scored by domain.
What question formats appear?
Five: multiple choice with one correct answer among four, multiple response with two or more correct answers among five or more options, ordering with three to five responses placed in sequence, matching with three to seven prompts, and case studies where one scenario carries two or more separately scored questions.
Can I reschedule or cancel?
Yes, up to 24 hours before the appointment, through the Manage Pearson VUE exams button in your AWS Certification account. Each appointment can only be rescheduled twice. Cancelling inside the 24 hour window, or not appearing, forfeits the exam fee and the delivery provider charges the full delivery fee with no refund available.
How long should I study for it?
Plan 70 to 100 hours over 10 to 12 weeks if you are new to AI and ML on AWS. Candidates who already hold an AWS certification and work with Amazon Bedrock day to day typically need far less, because AWS skips the foundational cloud courses in the prep plan for anyone holding Cloud Practitioner or an associate-level certification.
Does the exam test specific AWS service pricing figures?
No. The exam guide asks candidates to understand cost tradeoffs, for example token-based pricing against provisioned throughput, or RAG against fine-tuning, rather than to recall dollar amounts for specific instance types or model invocations.
Pass AWS Certified AI Practitioner (AIF-C01) — Guaranteed
94% pass rate on first attempt
One-time • Lifetime access
One-time payment