AWS Certification Path: How to Pick Your Route

The AWS certification path has four tiers: Foundational, Associate, Professional, and Specialty. Most people should start with Cloud Practitioner (CLF-C02) if they're new to cloud, then move to an Associate exam like Solutions Architect. If you already work in tech, you can skip the Foundational level and go straight to an Associate. There's no required order and no exam locks you out of the next one, so the "path" is really about picking the route that matches your job, not climbing a ladder in sequence.

Where you should start comes down to your background, not to any required sequence.

The four tiers at a glance

AWS groups its exams into four levels. Foundational proves you understand cloud basics. Associate exams prove you can do a specific job. Professional exams go deeper into architecture and operations. Specialty exams cover one narrow domain in depth.

Every one of these is valid for 3 years, and none of them list a hard prerequisite. AWS recommends experience for the higher tiers, but it doesn't gate the exams. You could sit the Professional Solutions Architect exam tomorrow with no other AWS cert. That's a bad idea for most people, but the option exists.

Where to start by background

The right starting point depends on what you already do for work.

If you're new to cloud, non-technical, or changing careers, start with Cloud Practitioner (CLF-C02). It's the cheapest exam at $100, it has no prerequisites, and its domains (Cloud Concepts, Security and Compliance, Cloud Technology and Services, Billing, Pricing, and Support) give you the vocabulary the rest of the path assumes you know. If your goal is an AI-adjacent role rather than infrastructure, AI Practitioner (AIF-C01) sits at the same Foundational level and the same $100 price, covering AI/ML and generative AI fundamentals.

Working developers can skip Foundational. Go to Developer - Associate (DVA-C02) if you write application code, since its biggest domain is Development with AWS Services at 32 percent. If you lean more toward design than code, Solutions Architect - Associate (SAA-C03) is the more popular and more broadly recognized choice.

CloudOps Engineer - Associate (SOA-C03) is built for sysadmins and ops people. Its domains are weighted toward Monitoring, Reliability, and Deployment and Automation, which maps to day-to-day operations work.

For a data or ML background, Data Engineer - Associate (DEA-C01) covers ingestion, transformation, and data stores. Machine Learning Engineer - Associate (MLA-C01) is aimed at people with hands-on SageMaker experience. Both are Associate-level at $150.

If you're genuinely unsure, default to the Cloud Practitioner to Solutions Architect route. It's the most common path and the most widely understood by hiring managers. We compared the two directly in Cloud Practitioner vs Solutions Architect if you want help choosing the entry point.

The Foundational tier

Cloud Practitioner (CLF-C02) is the front door. The exam is 90 minutes, 65 questions with 50 scored, and you need 700 out of 1000 to pass. It's designed to confirm you understand what the cloud is, how AWS bills for it, and the core services, without expecting you to build anything. For a full breakdown, see the AWS Cloud Practitioner study guide.

AI Practitioner (AIF-C01) is the newer Foundational option. Same $100 cost and 700/1000 passing score, but the content shifts to AI and generative AI. AWS recommends some exposure to AI/ML on AWS first, and points newcomers toward Cloud Practitioner basics before this one. The AI Practitioner study guide walks through the domains.

A foundational cert won't make you a senior engineer, but it does open doors. For the Cloud Practitioner specifically, one industry roundup counted 32,000-plus job postings referencing the cert with a year-over-year growth of about 22 percent (studytech.ai, 2026). Entry-level cloud engineering pay has been cited around $101,337 per year in the US (datacamp.com, 2026), though figures vary widely by source and region and shouldn't be read as a guarantee.

The Associate tier

This is where most of the real hiring signal lives. Every Associate exam is $150, 130 minutes, and needs 720 out of 1000 to pass.

Solutions Architect - Associate (SAA-C03) is the most popular AWS cert overall. Its domains are Design Secure Architectures (30 percent), Design Resilient Architectures (26 percent), Design High-Performing Architectures (24 percent), and Design Cost-Optimized Architectures (20 percent). If you only ever take one AWS exam, this is usually the one. The SAA-C03 study guide covers it in depth, and if you're wondering about difficulty, is the AWS SAA exam hard? is an honest take. The exact pass mark and what it means is in AWS SAA-C03 passing score.

Developer - Associate (DVA-C02) is the build-focused sibling. AWS recommends a year of hands-on experience and proficiency in at least one programming language. See the Developer Associate study guide for the details.

CloudOps Engineer - Associate (SOA-C03) covers running workloads, not designing them. Its content spans monitoring, reliability, automation, networking, and security. The SysOps certification guide covers what to expect.

Data Engineer (DEA-C01) and Machine Learning Engineer (MLA-C01) round out the Associate tier for data and ML people. MLA-C01 assumes at least a year of hands-on SageMaker work, so it's not a first cert for most.

The Professional tier

Professional exams are a real step up. They're $300, 180 minutes, and need 750 out of 1000 to pass.

Solutions Architect - Professional (SAP-C02) is the natural sequel to SAA. AWS recommends 2 or more years of hands-on AWS experience and suggests holding the Associate version first. The exam leans into organizational complexity and migration at scale.

DevOps Engineer - Professional (DOP-C02) is for people automating deployments and managing infrastructure as code. AWS recommends 2 or more years operating AWS environments plus scripting experience. Its domains run from SDLC Automation to Incident and Event Response.

Generative AI Developer - Professional (AIP-C01) is the newest Professional exam, focused on building production GenAI applications. AWS recommends roughly 2 years building on AWS plus a year of hands-on GenAI work. There's no certification prerequisite, but the recommended experience tells you who it's for.

Don't rush these. A Professional exam without the underlying experience tends to end in a fail and a wasted $300.

The Specialty tier

Specialty exams go deep on one domain. Both current ones are $300 and 170 minutes.

Security - Specialty (SCS-C03) needs 750 out of 1000 and covers detection, incident response, infrastructure security, IAM, data protection, and governance. AWS recommends 3 to 5 years securing cloud solutions. The Security Specialty guide breaks it down.

Advanced Networking - Specialty (ANS-C01) needs 700 out of 1000 and assumes serious networking depth: AWS recommends 5-plus years of networking with 2-plus years in cloud and hybrid networking. This is the exam for people who already live in VPCs, Transit Gateways, and Direct Connect.

Specialty certs aren't a tier you climb to for prestige. You take one because your job is that domain.

How long the whole path takes

There's no single answer, because it depends on your starting point and how much you study each week.

Cloud Practitioner is reachable in a few weeks of part-time study for most beginners. An Associate exam usually takes one to three months of consistent prep, more if you're learning the underlying concepts from scratch. Professional exams are the long pole; with the recommended 2-plus years of experience already in hand, expect a couple of months of focused study, and much longer if you're trying to substitute study for experience.

You don't need to collect all twelve. Most working professionals stop at one Associate, or one Associate plus one Professional or Specialty that matches their role. The path is a menu, not a checklist.

Your next step

If you're new, book Cloud Practitioner and start there. If you already work in tech, pick the Associate that matches your job and go straight to it. Either way, the fastest way to know whether you're ready is to work through realistic, exam-style questions until the patterns click.

You can practice with realistic exam questions for any tier, or jump straight to the Cloud Practitioner practice set to start the path.