AWS AI Practitioner: The $100 Way Into AWS’s AI Ladder

AWS Certified AI Practitioner is code AIF-C01, costs $100, runs 90 minutes and asks 65 questions. It is a foundational exam, priced the same as Cloud Practitioner, and it is the cheapest credential in an AI family that now runs to four exams, three of which come to $550 in fees between them.

Where it sits on the AWS certification ladder matters more than usual here, because AWS has been rebuilding this corner of its catalogue all year.

What the exam covers

AWS describes the audience as people who want to validate “foundational knowledge of AI/ML concepts and generative AI on AWS” — which in practice means vocabulary, service awareness, responsible-AI concepts and knowing which AWS product does what, rather than building anything.

That makes it a conversation exam rather than an engineering one. It suits product managers, analysts, consultants, support staff and developers who are adjacent to AI work and need to be credible in the room, which is a large and honest use case.

What AWS suggests you do first

The exam page is helpful about sequence: people new to IT are pointed at AWS Cloud Practitioner Essentials or AWS Technical Essentials before attempting this exam. That is a recommendation, not a requirement — AWS names no prerequisite certification for any of its exams.

If you have never used AWS at all, the honest order is to spend the $100 on the Cloud Practitioner exam first, because AI Practitioner assumes you already know what the platform is. If you know AWS and not AI, start here.

The four AI credentials, and what separates them

AWS now sells AI-related certifications at four points, and the differences are about job, not difficulty ranking.

• AI Practitioner (AIF-C01), $100, 90 minutes, 65 questions. Foundational. Concepts and services.
• Machine Learning Engineer – Associate (MLA-C01), $150, 130 minutes, 65 questions. For people who build and operate ML systems. Its replacement, MLA-C02, is in beta at $75 for 170 minutes and 85 questions.
• Generative AI Developer – Professional (AIP-C01), $300, 180 minutes, 75 questions. Professional level, for developers shipping generative-AI systems on AWS.
• AI Business Strategist (AIB-C01), beta at $50 with a standard price AWS gives as 100 USD, 170 minutes, 85 questions. For people making the business case rather than the system.

The machine-learning associate exam is mid-transition, and the dates are published: MLA-C01 retires in English on 28 September 2026, beta delivery of MLA-C02 begins 29 September 2026, and MLA-C02 becomes generally available in all languages on 14 January 2027. AWS’s September 2026 update describes the new version as leaning into generative AI, foundation models and agentic workflows.

What happened to the old ML Specialty exam

It closed. AWS’s own page for the Machine Learning – Specialty certification states “The last day to take this exam is March 31, 2026”, and adds that holders keep an active certification for three years from the date they earned it.

That is worth knowing if you are working from a study plan written before 2026 — a chunk of the AWS ML advice online still points at an exam that can no longer be booked.

Preparing without spending

For a $100 foundational exam, the free tier is the whole answer for most people: AWS’s Official Practice Question Sets, the two-hour Exam Prep Course, AWS Cloud Quest, and the 900-plus free self-paced digital courses on its training platform. None of it pays this site anything.

If you want more repetition, a month of Skill Builder at $29 buys the Official Practice Exam for this test, and that is the version of the subscription decision we would defend. A year at $449 for one foundational exam is not. The free tier against the paid one, product by product compares them properly. Third-party question banks such as Whizlabs also sell AI-exam sets cheaply — a partner link, tagged sponsored and nofollow, which may earn Stackrung a commission, and strictly a supplement to AWS’s own sets.

Is a foundational AI certification taken seriously?

It is taken as what it is. A foundational credential tells a reader that you have learned the vocabulary and can hold a conversation about AWS’s AI services without being sold to. It does not claim you can build a retrieval pipeline, and nobody reading your profile will think it does.

The failure mode is spending $300 on the professional generative-AI exam because the title sounds better. That exam is written for developers with two or more years of cloud experience shipping production systems, and the price of finding out otherwise is a full-fee retake after 14 days.

If your work is data rather than AI, the more useful adjacent credential is probably the Data Engineer – Associate exam, because the pipelines come before the models.

The three-year clock on a fast-moving subject

Like every AWS certification, AI Practitioner expires three years after you earn it, and unlike Cloud Practitioner there is no free recertification route for it. Renewal means retaking the current version, passing a higher-level exam, or using the paid Skill Builder maintenance route inside the last 90 days.

In a subject changing this quickly, that clock is less of an annoyance than it sounds. The AIF-C01 exam you sit today describes services that will have been renamed, merged or superseded well inside three years — AWS’s own September 2026 update rewrote the machine-learning associate exam around foundation models and agentic workflows, concepts that were marginal when the previous version was written.

The practical consequence is a planning one. If AI is where you are heading, treat this $100 exam as a dated snapshot rather than a durable credential, and expect the renewal to be a genuinely different exam rather than a formality. How each renewal route works, and what it costs sets out the options, including the 50% voucher AWS puts in your certification account when you recertify or upgrade.

What it looks like on a profile

A foundational certification reads as what it is: evidence you learned the vocabulary deliberately, on a date. That is worth something to a recruiter screening on keywords and very little to an interviewer asking what you have built.

The people it serves best are the ones whose job is to make decisions about AI rather than implement it — the product manager who needs to challenge a vendor’s claims, the consultant scoping a project, the analyst asked whether a use case is realistic. For them, $100 and 90 minutes buys a genuinely useful map. For a developer intending to build, it is a stepping stone to pass through quickly rather than a destination.

Booking and the retake rule

AWS delivers the exam through Pearson VUE, at a test centre or online proctored. You can reschedule twice before having to cancel and rebook, and final results reach your AWS Certification Account within five business days of the exam closing.

Fail, and the rule is the same as at every other level: 14 calendar days, then the full $100 again, with no limit on attempts. At foundational prices that is a survivable mistake, which is part of why this rung is a reasonable place to learn how AWS’s testing machinery feels.