AWS has two foundational certifications, and both are popular first steps. Cloud Practitioner proves you understand the AWS Cloud as a whole. AI Practitioner proves you understand AI, machine learning and generative AI, and how AWS delivers them. If you can only fit one into the next few months, the right choice depends on your role and on how much AWS you already know.
The two exams side by side
| Fact | Cloud Practitioner | AI Practitioner |
|---|---|---|
| Exam code | CLF-C02 | AIF-C01 |
| Questions | 65 (50 scored, 15 unscored) | 65 (50 scored, 15 unscored) |
| Time | 90 minutes | 90 minutes |
| Pass mark | 700 on a scale of 100 to 1,000 | 700 on a scale of 100 to 1,000 |
| Question types | Multiple choice, multiple response | Multiple choice, multiple response, ordering, matching |
| Fee | 100 USD | 100 USD |
| Validity | 3 years | 3 years |
| Delivery | Pearson VUE test centre or online | Pearson VUE test centre or online |
We checked both certification pages and exam guides on aws.amazon.com and docs.aws.amazon.com on 19 November 2026. Neither page had announced a replacement exam code. AWS has said the German and Italian versions of AIF-C01 retire after 15 October 2026, and the German and Italian versions of CLF-C02 after 31 December 2026, so check the language list if you do not plan to sit in English.
What each exam covers
Cloud Practitioner (CLF-C02)
| Domain | Weight |
|---|---|
| Cloud Concepts | 24% |
| Security and Compliance | 30% |
| Cloud Technology and Services | 34% |
| Billing, Pricing, and Support | 12% |
This is a broad survey. You need to explain the value of the cloud, the shared responsibility model, the Well-Architected Framework, global infrastructure, core compute, storage, database and networking services, and how AWS charges and supports customers. Coding, architecture design and troubleshooting are out of scope.
AI Practitioner (AIF-C01)
| Domain | Weight |
|---|---|
| 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 Solutions | 14% |
This one goes deep on a single theme. You need the vocabulary of machine learning, how foundation models and large language models behave, prompt engineering, retrieval augmented generation, fine tuning, model evaluation, responsible AI and the security of AI systems. Amazon Bedrock and Amazon SageMaker AI sit at the centre. The guide states clearly that building models, writing code and doing statistics are out of scope.
Where the content overlaps
The overlap is real but smaller than people expect. The AI Practitioner guide says the target candidate should already be familiar with core AWS services such as EC2, S3, Lambda, Bedrock and SageMaker AI, plus the shared responsibility model, IAM and AWS pricing models. That is roughly the Security and Compliance and Billing domains of Cloud Practitioner, plus part of Cloud Technology.
So Cloud Practitioner study helps AI Practitioner, but AI Practitioner study does not cover most of Cloud Practitioner. Networking, databases, support plans and the Well-Architected pillars barely appear in the AI exam.
If you take Cloud Practitioner first, you arrive at AI Practitioner with IAM, encryption, shared responsibility and pricing already learnt. That covers a useful slice of the AI exam's security domain and frees your time for the generative AI content.
Career fit
Pick Cloud Practitioner first if you:
- are moving into a cloud support, operations, sales or project role
- plan to go on to Solutions Architect, Developer or SysOps Associate
- have never used AWS and want a structured tour
Pick AI Practitioner first if you:
- work in product, business analysis, marketing or leadership and are being asked to judge AI proposals
- already know AWS basics from work
- plan to move towards the Machine Learning Engineer Associate track
One detail worth knowing: AWS states that earning the Machine Learning Engineer Associate certification automatically recertifies AI Practitioner. That makes AI Practitioner a natural first step on the machine learning path.
Compensatory scoring and what it means for you
Both exams use compensatory scoring. You pass on the overall scaled score, not domain by domain, so a strong domain can make up for a weaker one. Neither exam penalises guessing, and unanswered questions are simply marked wrong, so always answer everything.
There is one important difference. The AI Practitioner guide states that for multiple response, ordering and matching questions you must get every part right to earn the mark. There is no partial credit. A matching question with five pairs and one wrong pair is worth nothing.
On AIF-C01 ordering and matching items, settle the pairs you are certain of first, then use elimination on the rest. Because the item is all or nothing, one careful extra minute here is often worth more than rushing to the next question.
Generative AI traps on AI Practitioner
These patterns catch people who have used chat tools but have not studied the concepts:
- Fine tuning versus RAG. If the scenario needs answers from frequently changing company documents, retrieval augmented generation is usually the better fit. Fine tuning changes model behaviour or style and is costlier to refresh.
- Temperature. Lower temperature makes output more predictable, higher makes it more varied. It does not make a model more accurate.
- Hallucination fixes. Grounding the model in trusted data and adding guardrails address hallucination. Simply choosing a larger model does not guarantee it.
- Pricing. Generative AI services are often priced per token, for both input and output. Long prompts cost money too.
- Responsible AI terms. Know the difference between bias, fairness, explainability, transparency and robustness. Questions use them precisely.
- Service names. AWS renames and adds services often. Learn what each service does rather than memorising marketing names.
Answer options that promise to "eliminate" bias or hallucination entirely are almost always wrong. The exam favours options that reduce or monitor risk.
A simple decision guide
Ask yourself three questions in order.
First, have you used AWS in a job or project? If not, start with Cloud Practitioner. The AI exam assumes basic AWS familiarity and you will spend extra time learning it anyway.
Second, if you do know AWS basics, is your next career move about AI? If yes, take AI Practitioner first. If no, take Cloud Practitioner and keep AI for later.
Third, are you aiming at a technical associate certification soon? If it is Solutions Architect, Developer or SysOps, Cloud Practitioner is the better warm up. If it is Machine Learning Engineer or Data Engineer, AI Practitioner gives you more useful groundwork.
A combined study path
Many people end up taking both. This order works well for a newcomer with about five hours a week:
- Weeks 1 to 4: Cloud Practitioner study. Cover all four domains, open a Free Tier account and click through IAM, S3, EC2 and the billing console.
- Week 5: Cloud Practitioner mocks and exam. Two timed mocks, review every miss, then sit the exam.
- Weeks 6 to 7: AI and ML fundamentals. Learning types, the ML lifecycle, inference options and evaluation metrics.
- Weeks 8 to 9: generative AI and foundation models. Tokens, embeddings, prompts, RAG, fine tuning, agents, Bedrock and its guardrails.
- Week 10: responsible AI and security. Then two timed AIF-C01 mocks and the exam.
If you already know AWS, skip the first five weeks and use what you know as the base for AI Practitioner. MockDen has study material and timed mocks for both, at AI Practitioner and Cloud Practitioner, so you can practise each in the same format as the real exam.
Try it for yourself
The quickest way to decide is to answer a few questions from each exam and see which feels closer. Try the free practice on MockDen for both certifications, compare your scores, and start with the one that fits your role. MockDen is an independent study site and is not affiliated with AWS.
Last checked against official sources on 19 November 2026.
Frequently asked questions
Is AI Practitioner harder than Cloud Practitioner?
Most people find it harder if they are new to AI, because it adds machine learning and generative AI concepts on top of basic AWS knowledge. The format, length and pass mark are the same.
Do I need Cloud Practitioner before AI Practitioner?
No. Neither exam has a prerequisite. AWS recommends familiarity with core services such as EC2, S3, Lambda and IAM, which Cloud Practitioner study covers well.
Do I need to code for AI Practitioner?
No. The exam guide lists coding AI models, feature engineering and hyperparameter tuning as out of scope.
How much do the two exams cost?
AWS lists each at 100 USD. Taxes and currency conversion can change the final amount.
- AWS, AWS Certified Cloud Practitioner
- AWS, AWS Certified AI Practitioner
- AWS, CLF-C02 exam guide
- AWS, AIF-C01 exam guide
- AWS, Certification policies after testing
MockDen is independent and not affiliated with the exam owners. Always confirm fees and dates on the official page before booking.