A clean way to burn $300 is to book the Solutions Architect Professional exam because the title sounds senior, pass it, then list it on a resume for a role that wanted three years of hands-on AWS you haven’t done yet. The credential is real. The signal it sends to that hiring manager is weaker than you’d hope, because anyone who reads senior cloud resumes can tell a cert apart from a deploy history.
So the honest answer to whether an AWS cert is worth it depends almost entirely on where you’re standing. For someone with no cloud job trying to get past a resume filter, a cert can be the thing that earns a first call. For a staff engineer who already runs production on AWS, it’s mostly a line a recruiter skims past. Same piece of paper, very different value.
What you actually pay
AWS prices exams by tier, and the numbers have held steady through 2026. The foundational exams (Cloud Practitioner and the newer AI Practitioner) run $100. Associate exams, including Solutions Architect Associate, Developer, SysOps, Data Engineer, and Machine Learning Associate, are $150. Professional exams (Solutions Architect Professional, DevOps Engineer Professional) and the Specialty exams sit at $300. Local tax gets added on top depending on where you test, so the charge on your card may be a bit higher than the sticker.
Two things soften the cost. Pass any exam and AWS hands you a 50% discount voucher good for one attempt within the next year, so a planned two-cert run is cheaper than it looks on paper. And the practice questions plus a good chunk of the official prep content are free through AWS Skill Builder. The exam fee is rarely the expensive part. Your time is. Budget against the table below, not against the $150.
One caveat on pricing and the lineup: AWS revises both. Regional pricing varies, taxes vary, and AWS sunsets older exams on a schedule (several legacy Specialty exams retire on June 30, 2026). Before you book anything, confirm the current fee and whether the exam still exists at aws.amazon.com/certification.
| Certification | Level | Exam fee (USD) | Best fit | Rough prep time |
|---|---|---|---|---|
| Cloud Practitioner | Foundational | $100 | Career switchers, PMs and sales near cloud teams | 20-40 hours |
| AI Practitioner | Foundational | $100 | Same audience wanting AI vocabulary, not ML engineers | 20-40 hours |
| Solutions Architect Associate | Associate | $150 | The default first real cert; broad service coverage | 60-120 hours |
| Data Engineer Associate | Associate | $150 | Analytics and data-platform roles | 80-140 hours |
| Security Specialty | Specialty | $300 | Security or platform engineers already in the work | 80-160 hours |
| Solutions Architect Professional | Professional | $300 | Engineers with real AWS production experience | 120-200 hours |
Those prep hours are a loose guess and depend hard on what you already do at work. Someone who deploys to ECS every day can pass the Developer Associate on a weekend of review. Someone who has never opened the console needs weeks for the same exam.
Which cert matches which goal
If you have no cloud job and want one, the Solutions Architect Associate is the cert that does the most work. It’s broad enough to push you through the services that show up in real interviews (VPC, IAM, S3, the difference between an ALB and an NLB, when to reach for a queue), and it’s recognized widely enough that a screener knows what it means. The Cloud Practitioner is a gentler on-ramp and fine if you’re a project manager or in sales engineering, but on its own it rarely moves an engineering resume. Hiring managers read it as “started learning,” which is honest but not a hiring signal.
If you already have a cloud job, the math changes. Pick the cert that matches what you touch. A data engineer gets more out of the Data Engineer Associate than a generic architecture cert, because the questions map to Glue, Redshift, and the streaming patterns you can actually talk about in an interview. A security engineer who passes the Security Specialty walks away with a credential that lines up with the day job, which is the only version of a cert that survives a real follow-up question.
The Professional exams are a different animal. Solutions Architect Professional and DevOps Engineer Professional assume you’ve designed and run nontrivial systems on AWS. You can pass them by grinding practice questions without that background, and some people do, but the gap shows the moment a system-design interviewer pushes past the memorized answer. These pay off most when you already have the experience and want the credential to match it, often because a consulting firm or an AWS partner-tier requirement asks for it.
The 2026 AI cert wave, and what to ignore
AWS leaned hard into AI naming over the last year. The AI Practitioner (foundational) and the Generative AI Developer track are the headline additions, and the marketing around them is loud. Stay skeptical in proportion to the noise. The AI Practitioner is a foundational exam: it teaches vocabulary, foundation-model basics, and where Bedrock fits. That’s genuinely useful for a PM or a solutions architect who needs to speak the language, and close to worthless as a signal for an ML engineer, who will be asked about retrieval quality, eval design, and serving cost, none of which a foundational cert touches.
If you’re aiming at actual GenAI engineering roles, the badge is not what gets you hired. A small project you can demo (a RAG system with a real evaluation set, a fine-tune with before-and-after numbers) outscores any 2026 AWS AI cert in an interview, because it answers the question the cert can’t: can you make the thing work and measure whether it got better.
Who it genuinely helps
Certs do real work in a few specific spots. Career switchers and bootcamp grads use them to clear automated resume filters and to prove they were serious enough to study, and for that audience an Associate cert is a reasonable bet. Contractors and consultants benefit because clients and AWS’s own partner program count certified staff, so a cert can directly affect whether your shop wins work or reaches a partner tier. And plenty of companies pay a bonus or a raise tied to certs, in which case the math is easy: if your employer hands you a $1,000 bonus and a paid exam voucher, take it.
The group that gets the least are senior engineers with a visible track record. If your commits, your design docs, and your last three years of work already say you know AWS, a cert sits below your existing signal. It won’t hurt you, but the study time almost always returns more if you spend it on a side project, an open-source contribution, or interview prep aimed at the specific loop you’re walking into.
About those salary numbers
You’ll see claims that certified people earn 20 to 30 percent more, usually citing vendor surveys. Treat those as correlation, not a raise you can book. The people who collect AWS certs also tend to work at companies that pay for cloud skills, switch jobs to chase comp, and already have the kind of resume that lands cert-friendly roles. The cert is tangled up with all of that. It’s a contributing factor for someone breaking in, and a rounding error for someone already paid well to run cloud infrastructure. When a recruiter or a training platform quotes you a precise salary lift for a specific badge, ask where the number comes from and who paid for the survey.
The part that holds up: a well-chosen cert lowers the friction of getting interviews when your resume is thin, and the understanding you build studying for an Associate exam is worth more than the badge stapled to it. Stop at two or three that map to your path. Collecting all thirteen is a hobby, not a strategy, and the people reading your resume can tell the difference between someone who learned the platform and someone who learned the exam.
Useful next steps: