Your role determines your AWS AI certification path. AWS offers four AI-related certifications in 2026: AI Practitioner (AIF-C01, $100), Machine Learning Engineer Associate (MLA-C01 or MLA-C02, $150), Machine Learning Specialty (MLS-C01, $300), and Generative AI Developer Professional (AIP-C01, $300). None of them is called “AWS AI Architect certification” or “AWS AI Professional certification.” Those are search terms, not official AWS certification names. If you searched for either, you likely want AIP-C01 or MLS-C01. This article maps the official AWS AI certification path to the right target based on your actual job function.
All three exams require deliberate preparation. AWS does not publish official pass rates. No AWS AI certification is passable by casual familiarity. Certifications are valid for three years. Retakes require a 14-day wait and the full exam fee each time.
What Most Articles on This Topic Get Wrong
Most articles on this topic stop at AIF-C01 and MLS-C01. What they miss:
- The September 2026 MLA-C01 retirement deadline and what it means for your timeline
- AIP-C01 as a distinct Professional-level exam with a separate candidate profile from MLS-C01
- How the 50% recertification voucher cuts your total path cost by $75 to $150
- What “AWS AI certification free” and “AWS AI certification path PDF” actually refer to (they are not standalone AWS products)
- That passing MLA-C01 or MLA-C02 automatically recertifies an existing AIF-C01 credential
This article covers it all using only official AWS sources.
The Search Terms That Are Not Official AWS Certifications
Several high-volume search terms do not map to a single official AWS certification page. Here is what each one actually means:
| Search Term | What It Really Refers To |
| AWS AI Professional certification | AWS Certified Generative AI Developer Professional (AIP-C01) |
| AWS AI Architect certification | No official AWS certification by this name. Closest option: AWS Certified Solutions Architect + AIP-C01 |
| AWS AI certification path PDF | The official exam guide PDF for each cert, available free at aws.amazon.com/certification |
| AWS AI certification free | Free prep via AWS Skill Builder. The exam itself costs $100 to $300. There is no free AWS AI certification exam |
| AWS Gen AI certification path | AIF-C01 for concepts, AIP-C01 for production generative AI development |
AWS AI Certifications in 2026: Official Comparison Table
| Certification | Level | Exam Code | Cost | Questions | Duration | Passing Score |
| AWS Certified AI Practitioner | Foundational | AIF-C01 | $100 | 65 | 90 min | 700/1000 |
| AWS Certified Machine Learning Engineer Associate | Associate | MLA-C01 (retiring Sept 28, 2026) | $150 | 65 | 130 min | 720/1000 |
| AWS Certified Machine Learning Specialty | Specialty | MLS-C01 | $300 | 65 | 180 min | 750/1000 |
| AWS Certified Generative AI Developer Professional | Professional | AIP-C01 | $300 | 75 | 180 min | Not published |
All costs are in USD before applicable taxes. AWS Certifications include a 50% discount voucher after passing any exam, applicable to your next purchase.
Thinking of pursuing the AWS certifications? Explore: Is AWS Certification Hard To Pass? Get Deep Insights
Your AWS AI Certification Path: Choose Your Track
Picking the wrong AWS AI exam is a real and expensive mistake. Without a clear decision framework, candidates spend $150 to $300 on an exam that does not match their role or experience level.
Business Analyst / Product Manager / IT Manager / Executive
AIF-C01 → (Optional) AIP-C01 if working directly with Bedrock teams
Skip: MLA-C01, MLS-C01
Software Developer / Backend Engineer
AIF-C01 (optional) → MLA-C01 → MLS-C01 or AIP-C01
Skip: Cloud Practitioner (redundant if you already build on AWS)
Data Engineer / MLOps Engineer / DevOps Engineer
MLA-C01 (before Sept 28) or MLA-C02 beta (opens Sept 1) → MLS-C01
Skip: AIF-C01 unless you need the automatic recertification benefit
Developer Building Generative AI Apps with Bedrock
AIF-C01 → AIP-C01
Skip: MLS-C01 (wrong domain for this role)
Career Changer with No Cloud Background
AIF-C01 → MLA-C01 (after 6 to 12 months of hands-on experience)
Skip: AIP-C01 or MLS-C01 as a first exam
Track 1: Non-Technical Path (AIF-C01)
AIF-C01 is the AWS Certified AI Practitioner exam. It validates knowledge of AI/ML concepts, AWS AI services, generative AI basics, responsible AI, and governance. You do not need to write code or deploy a model to pass.
Who this is for:
- Business analysts evaluating AI tools for procurement decisions
- Product managers defining requirements for AI-powered features
- Executives approving AI investment and overseeing governance
- IT managers advising on AWS AI service adoption
- Career changers building a credential before targeting associate-level exams
Exam specifics:
- Cost: $100
- Duration: 90 minutes
- Format: 65 questions, multiple choice and multiple response
- Passing score: 700 out of 1000
- Valid for 3 years
- Testing: Pearson VUE testing center or online proctored exam
Domain weights:
- Fundamentals of AI and ML: 20%
- Fundamentals of generative AI: 24%
- Applications of foundation models: 28%
- Guidelines for responsible AI: 14%
- Security, compliance, and governance for AI: 14%
Applications of foundation models carry the highest weight. Spend proportionally more time there, not on edge-case theory.
What AIF-C01 does not do: It does not qualify you to build ML pipelines, deploy SageMaker endpoints, or architect Bedrock solutions. If your job involves any of those tasks, this exam is a starting point, not a destination.
Track 2: Engineering Path AIF-C01 → MLA-C01 / MLA-C02 → MLS-C01
This path is for engineers who build, train, deploy, and monitor machine learning workloads on AWS.
Critical 2026 Update: MLA-C01 Retires September 28
AWS confirmed the following timeline for the Machine Learning Engineer Associate exam:
- MLA-C01 (English) last exam date: September 28, 2026
- MLA-C02 beta registration opens: September 1, 2026
- MLA-C02 available in English only at beta launch; Korean, Japanese, and Simplified Chinese added later
If you are currently preparing for MLA-C01, sitting before September 28 locks in your result under the existing exam structure. Waiting for MLA-C02 means less community preparation material is available when you sit, since beta exams have limited third-party coverage.
This is the most time-sensitive decision in the AWS AI certification path right now.
MLA-C01 exam specifics:
- Cost: $150
- Duration: 130 minutes
- Format: 65 questions
- Passing score: 720 out of 1000
- AWS recommended experience: At least 1 year with Amazon SageMaker and related ML engineering services
What this exam actually tests:
Candidates without hands-on SageMaker experience report finding the exam significantly harder than those with it. AWS uses compensatory scoring: you do not need a passing sub-score in each domain, only an overall passing score. But domains like data preparation, model deployment, and CI/CD for ML pipelines each appear in scenario-based questions that require applied judgment rather than mere recall.
Core skills tested:
- Data ingestion and preparation with SageMaker Data Wrangler and AWS Glue
- Model training and hyperparameter tuning in SageMaker
- Deploying models to SageMaker endpoints with auto scaling
- MLOps pipelines and CI/CD for ML workflows
- Model monitoring and post-deployment performance tracking
Recertification benefit: Passing MLA-C01 or MLA-C02 automatically recertifies an existing AIF-C01 credential. This makes the associate exam a high-value target for anyone who already holds AIF-C01 and is approaching their three-year renewal window. You can review AWS recertification options for full details.\
Explore where AWS certification takes you: AWS certifications leading you to 9 great job positions in 2026
AWS Certified Machine Learning Specialty (MLS-C01)
MLS-C01 is the deepest AWS AI certification available. It targets experienced professionals who design, implement, and maintain production ML systems, including distributed training, algorithm selection, large-scale feature engineering, and model optimization.
Exam specifics:
- Cost: $300
- Duration: 180 minutes
- Format: 65 questions
- Passing score: 750 out of 1000
- AWS recommended experience: 2 or more years developing, architecting, and running ML workloads on AWS
Domain breakdown:
- Data engineering: 20%
- Exploratory data analysis: 24%
- Modeling: 36%
- ML implementation and operations: 20%
Modeling carries 36% of the exam weight. If you are preparing for MLS-C01 and do not have a strong background in algorithm selection, deep learning, and hyperparameter optimization, that section alone will likely determine your result.
Track 3: Generative AI Developer Path AIF-C01 → AIP-C01
AIP-C01 is the AWS Certified Generative AI Developer Professional exam. Despite the name, it is not a general “AWS AI Professional certification.” It specifically validates the ability to build and deploy production-grade generative AI solutions using Amazon Bedrock.
This exam is not for beginners. AWS recommends 2 or more years of production application experience on AWS or with open-source technologies, plus 1 year of hands-on experience implementing generative AI.
Who this is for:
- Developers who build RAG pipelines, multi-agent systems, or fine-tuned foundation model applications
- Engineers already working with Amazon Bedrock in a production or pre-production context
- Technical leads overseeing generative AI adoption with security and cost efficiency requirements
Exam specifics:
- Cost: $300
- Duration: 180 minutes
- Format: 75 questions, multiple choice and multiple response
- Testing: Pearson VUE testing center or online proctored
Core topics tested:
- Foundation model selection and evaluation in Amazon Bedrock
- Prompt engineering and prompt optimization techniques
- RAG pipeline architecture and implementation
- Multi-agent system design
- Fine-tuning workflows
- Responsible AI in production generative AI contexts
- Security and cost optimization for Bedrock-based solutions
- How AIP-C01 differs from MLS-C01: MLS-C01 covers classical ML theory, algorithm selection, statistical analysis, and SageMaker in depth. AIP-C01 covers generative AI architecture, Bedrock, RAG, and production deployment of foundation model applications. Candidates with strong Bedrock experience but limited classical ML background typically find AIP-C01 more approachable than MLS-C01. They are not interchangeable certifications.
Delve into the details of: Why to Endeavor AWS Machine Learning Specialty Certification
AWS AI Certification Cost: Full Path Breakdown
Paying full price for every exam is the most expensive outcome. One voucher changes the math.
| Path | Exams | Full Price | With Voucher Applied |
| Non-technical only | AIF-C01 | $100 | N/A (first exam, no voucher yet) |
| Non-technical then engineering | AIF-C01 + MLA-C01 | $250 | $175 (50% off MLA-C01) |
| Full engineering path | AIF-C01 + MLA-C01 + MLS-C01 | $550 | $325 (voucher applied to MLS-C01) |
| Generative AI developer path | AIF-C01 + AIP-C01 | $400 | $250 (50% off AIP-C01) |
| Single retake | Any exam, full fee | +$100 to $300 | No retake discounts available |
The 50% voucher is earned by passing any AWS certification. It applies to the next exam you purchase. Applying to MLA-C01 saves $75. Applying to AIP-C01 or MLS-C01 saves $150.
Organizations supporting team certification can purchase AWS exam vouchers in bulk. AWS does not offer free retakes under any plan or program.
Delve into the details to explore what it takes to earn Generative AI developer: AWS Certified Generative AI Developer Certification Cost in 2026
“AWS AI Certification Free”: What This Actually Covers
Searching for “AWS AI certification free” returns results that conflate free prep resources with a free exam. These are not the same thing.
What is actually free:
- AWS Skill Builder free tier: exam readiness courses, Official Practice Question Sets, and learning paths for AIF-C01, MLA-C01, and AIP-C01
- Official exam guides in PDF format for every AWS certification (download directly from aws.amazon.com/certification)
- AWS Machine Learning Best Practices whitepaper and Amazon Bedrock User Guide
- AWS Builder Labs on AWS free tier accounts for SageMaker and Bedrock hands-on practice
- AWS Training Live on Twitch: live and on-demand sessions from AWS experts
What is not free:
- The exam itself. AIF-C01 costs $100. No AWS AI certification exam is offered at zero cost outside of rare promotional programs run for limited periods.
- AWS Skill Builder Individual subscription (required for some premium courses, not required for the free Official Practice Question Sets)
Reality check for AIF-C01 prep: Free Skill Builder resources are sufficient to pass AIF-C01 for most candidates with a structured study schedule. The official practice question set and exam guide, combined with 4 to 6 weeks of part-time study, is the baseline most prepared candidates use.
Reality check for MLA-C01 and AIP-C01 prep: Free resources provide the theory. They do not replace hands-on lab work. AWS free-tier accounts support SageMaker Studio and Bedrock API calls within limits, which is enough to practice the core workflows tested in both exams. Candidates who skip hands-on practice and rely only on video courses report the scenario-based questions as the hardest part of both exams.
“AWS AI Certification Learning Path” and “AWS AI Certification Path PDF”: What These Mean
Neither is a single AWS product. Here is what each search intent actually maps to:
AWS AI certification learning path: AWS Skill Builder organizes exam prep into structured plans called “Exam Prep Plans” for each certification. The four-step AIF-C01 Exam Prep Plan is available on Skill Builder. The MLA-C01 Exam Prep Plan is available at the same location. These are the closest thing to an official AWS AI certification learning path.
AWS AI certification path PDF: This refers to the official exam guide published by AWS for each certification. Each exam guide is a downloadable PDF that lists domains, sub-tasks, AWS services in scope, and services explicitly out of scope. The AIF-C01 and MLA-C01 exam guides are available for free from the official certification pages.
There is no single “AWS AI certification path PDF” that covers all AWS AI certifications in a single document. You download a separate exam guide for each exam you are targeting.
2026 Labor Market Context
Data scientist employment is projected to grow 34% from 2024 to 2034, according to the U.S. Bureau of Labor Statistics. Employment of software developers is projected to grow 17.9% from 2023 to 2033. Both rates are significantly above the 3.1% projected growth for all occupations over the same period.
The World Economic Forum’s Future of Jobs Report projects that demand for AI and ML specialists will grow 40% in the near term. At the same time, 70% of North American IT leaders report difficulty filling AI/ML roles.
An AWS-commissioned study found that employers are willing to pay 43% more for IT workers with demonstrated AI skills, 42% more for finance roles, and 41% more for business operations roles. The Wall Street Journal reported that 1 in 4 technology job postings now list AI skills as a requirement.
These numbers establish the supply gap. They do not guarantee that any specific certification will result in a pay increase or a job offer. Holding MLA-C01 or AIP-C01 removes you from the unverified majority and gives hiring managers a concrete, standardized signal. Combined with demonstrable project experience, that signal carries weight. Without portfolio evidence, it carries less.
Get insights on: Exploring Salary Trends and Career Worth of AWS Machine Learning Engineer Associate Certification
20 Types of Questions Should You Expect on the AI-103 Exam 2026
The following question types and scenarios are based on the official AI-103 skills measured as of April 16, 2026, and reflect the five exam domains published in the Microsoft Learn study guide. They are not real exam questions and have not been sourced from any unauthorized material. Instead, they represent the kinds of applied, scenario-driven questions you are likely to encounter, based on the technical competencies Microsoft explicitly tests. Use them to identify gaps in your preparation and to practice applied reasoning across generative AI, agentic solutions, computer vision, text analysis, and information extraction on Azure.
- Which Microsoft Foundry service should you choose when building a grounding pipeline that requires vector search over multimodal content?
- How do you configure a RAG pipeline in Azure to connect document retrieval directly to an agent tool?
- When deploying a large language model through Microsoft Foundry, which deployment option minimizes latency for high-throughput agent workloads?
- What steps are required to implement retrieval-augmented generation (RAG) in a Python-based Azure AI application using Foundry SDKs?
- How do you build a multi-agent orchestration solution in Azure that supports autonomous workflows with human-in-the-loop approval controls?
- Which Azure service is used to extract structured fields and layout information from scanned documents using OCR and multimodal pipelines?
- How do you configure safety filters and content moderation guardrails in a generative AI application deployed on Microsoft Foundry?
- What is the correct approach to implementing managed identity and keyless credentials for securing an Azure AI solution?
- How do you monitor model performance, grounding quality, and safety events in a deployed Azure AI agent?
- Which method should you use to configure semantic search, hybrid search, and vector search for a document grounding pipeline in Azure AI Search?
- How do you implement prompt engineering and chain-of-thought evaluation to optimize generation behavior in a Foundry-based application?
- What configuration is required to enable speech-to-text and text-to-speech modalities in an Azure AI agent interaction workflow?
- How do you build a multimodal understanding workflow that enables visual question answering grounded in image evidence?
- Which Azure Content Understanding pipeline mode should you use when processing video segments to extract visual characteristics?
- How do you detect and mitigate indirect prompt injection attacks embedded in images within a computer vision solution?
- What responsible AI instrumentation steps are required to implement trace logging, provenance metadata, and approval workflows in an agentic system?
- How do you configure an Azure AI Language solution to extract entities, detect sentiment, and produce structured JSON outputs using generative prompting?
- Which Foundry service should you select to translate speech into another language within an agentic interaction pipeline?
- How do you integrate a Foundry project with a CI/CD pipeline to manage continuous deployment of AI models and agent workflows?
- What approach should you use to manage quotas, rate limits, and cost footprints for model and agent workloads at scale in Azure?
Frequently Asked Questions
Which AWS AI certification should I get first?
Start with AIF-C01 if you are in a non-technical role or are new to AWS AI services. Start with MLA-C01 if you already have engineering experience and hands-on AWS exposure. Do not start with AIP-C01 or MLS-C01 as your first certification. Both require applied experience that the exam tests directly.
What is the “AWS AI Architect certification”?
There is no official AWS certification by that name. If you are looking to validate AI and cloud architecture skills together, the closest combination is AWS Certified Solutions Architect Associate plus AIP-C01. Some organizations use “AI Architect” as a job title, but AWS does not offer a single exam under that label.
What is the “AWS AI Professional certification”?
This search term refers to the AWS Certified Generative AI Developer Professional (AIP-C01). It is not a separate certification. AWS uses “Professional” as a certification level, not a standalone category. AIP-C01 is the only Professional-level AI exam AWS currently offers.
Is there a free AWS AI certification?
No exam is free. The AWS Certified AI Practitioner costs $100. Prep resources including official practice question sets, exam guides, and Skill Builder courses are available at no cost. AWS free-tier accounts provide access to SageMaker Studio and Amazon Bedrock for hands-on practice, subject to the service limits outlined on the AWS Free Tier page.
When does MLA-C01 retire?
The English-language MLA-C01 exam is available through September 28, 2026. MLA-C02 beta registration opens September 1, 2026. If you are mid-preparation, finishing before September 28 is the lower-risk option. MLA-C02 beta will have less community preparation material available at launch.
Do I need prior AWS certifications to take the MLA-C01?
No formal prerequisite exists. In practice, candidates without at least one year of hands-on SageMaker experience report the exam as significantly harder. Review the AWS Certification FAQs for full policy details on prerequisites and exam eligibility.
What happens if I fail an AWS AI certification exam?
You must wait 14 calendar days before retaking the same exam. Each retake requires the full exam fee. There is no cap on the number of attempts, and there is no free retake program.
Does passing MLA-C01 renew my AIF-C01 certification?
Yes. Passing MLA-C01 or MLA-C02 automatically recertifies an existing AWS Certified AI Practitioner credential. All AWS certifications are valid for three years from the date of passing.
How is AIP-C01 different from MLS-C01?
MLS-C01 tests classical ML theory, algorithm selection, feature engineering, data analysis, and depth in SageMaker. AIP-C01 tests generative AI architecture, Amazon Bedrock, RAG pipelines, prompt engineering, and production deployment of foundation model applications. They are not substitutes. Your target role determines which is relevant. The AIP-C01 exam guide outlines the full domain breakdown for direct comparison.
Can I prepare for AIF-C01 entirely with free resources?
Yes. Free Skill Builder resources, the official exam guide PDF, and AWS whitepapers are sufficient for most AIF-C01 candidates with a structured schedule. For MLA-C01 and AIP-C01, free resources cover the theory, but hands-on lab practice using an AWS free-tier account is necessary to handle the scenario-based questions.
