All Exam Questions
15+ Certification Practice Tests Available

Google Cloud Certification Exams: Tiers, Fees, Retakes and Renewal

Google Cloud certifications run from no-code leadership credentials to professional exams for architects, data engineers and now agentic AI builders. See the tiers, fees, retake rules and renewal options, and find the right starting point.

15+ Exams
25+ Practice Tests
1338+ Questions
4+ Categories
Google Cloud logo
Provider Overview

Google Cloud

Active
15+
Available Exams
25+
Practice Tests
1338+
Questions
4+
Categories

Google Cloud

Professional Machine Learning Engineer

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Google Cloud Certified – Professional Cloud Database Engineer

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Google Cloud Associate Cloud Engineer

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Professional Google Workspace Administrator

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Google Cloud Certified Associate Data Practitioner

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Google Cloud Certified – Professional Cloud Network Engineer

1 Practice Exams

50 Questions

Start Practicing

Google Cloud

Google Cloud Digital Leader

1 Practice Exams

50 Questions

Start Practicing

Who Google Cloud Certifies, and What for

Google’s certification program covers one thing: how well you work with Google Cloud products. That means Compute Engine, GKE, BigQuery, Cloud Run, the Gemini platform and the rest of the stack. It is not a general cloud-theory program. Nearly every exam is tied to a job role (engineer, architect, security engineer, database engineer), and Google’s own exam guides frame questions around business scenarios for that role You can also prepare for the AWS Certified Cloud Practitioner exam with our free practice questions.

The current catalog on Google’s certification page has 15 credentials: two foundational, three associate and ten professional. Google has also said its exams are being updated for the product announcements from Google Cloud Next ‘26, including the Gemini platform and the data and analytics stack. If you sit an exam, read the current exam guide first. Product names and coverage have shifted noticeably in the last year.

All exams are open to anyone aged 18 or older, and none of them has a formal prerequisite. The “recommended experience” lines on exam pages are advice, not gates.

Certification Tier Structure

Google uses three tiers. Unlike AWS, there is no separate Specialty tier. The specialized exams (security, networking, databases, machine learning) sit inside the Professional tier. There is also no Expert tier above Professional.

Tier

Exams listed today

Who it suits

Validity

Foundational

Cloud Digital Leader, Generative AI Leader

People who work alongside technical teams: managers, analysts, sales and product staff. No technical prerequisites.

3 years

Associate

Cloud Engineer, Google Workspace Administrator, Data Practitioner

Practitioners who deploy, monitor and maintain solutions. Google recommends 6+ months of hands-on Google Cloud experience for Cloud Engineer.

3 years

Professional

Agentic Architect, Cloud Architect, Cloud Database Engineer, Cloud Developer, Data Engineer, Cloud DevOps Engineer, Cloud Security Engineer, Cloud Network Engineer, Machine Learning Engineer, Security Operations Engineer

Experienced people who design, build and run production systems. Google recommends 3+ years of industry experience, including 1+ years on Google Cloud, for Cloud Architect.

2 years

Note that Workspace Administrator is about Google Workspace (Gmail, Drive, identity administration), not Google Cloud infrastructure. It shares the Associate tier but suits a different career track.

Which Certifications Matter Most Right Now

“In demand” is a judgment call, and Google publishes no ranking. What follows is informed opinion based on how often these credentials show up in cloud job descriptions and how central their subject matter is to Google Cloud’s pitch.

  • Professional Cloud Architect is the flagship. Case studies are part of the exam, and the questions reward judgment about trade-offs over memorized product facts.

  • Associate Cloud Engineer is the most common first technical credential and the usual stepping stone toward the Professional exams.

  • Professional Data Engineer and Professional Machine Learning Engineer lean on Google Cloud’s strongest territory, with BigQuery, Dataflow, Pub/Sub and the Gemini platform’s ML tooling.

  • Professional Cloud Security Engineer and Professional Cloud DevOps Engineer are the picks if your day job is IAM, policy and incident response or SRE and delivery pipelines.

  • Generative AI Leader is the newest foundational credential. It has no coding and targets business decision-makers.

  • Professional Agentic Architect is the newest professional exam and the one to watch. It is covered below.

What Is Different About the Agentic Architect Exam

This credential has two parts. Part 1 is a proctored multiple-choice exam delivered by Pearson. Passing it unlocks Part 2, hands-on labs in Google Skills that test whether you can actually build and deploy. The beta exam ran about 80 questions over three hours at a discounted $120 against a $200 retail price. Beta passers have until December 31, 2026 to finish the labs. Google’s page says general registration opens in mid-November 2026. The beta listing shows a one-year validity period, which differs from the two years Google’s FAQ gives for professional certifications, so check the final listing before you plan around it.

Exam Fees and Formats at a Glance

Fees below come from Google’s own exam pages and exclude local taxes.

Exam

Fee (USD)

Length

Format

Generative AI Leader

$99

90 minutes

Multiple choice, English

Cloud Digital Leader

$99

90 minutes

Multiple choice, English and Japanese

Associate Cloud Engineer

$125

2 hours

50–60 multiple choice and multiple select

Professional Cloud Architect

$200

2 hours

50–60 multiple choice and multiple select, with case studies

Professional Agentic Architect

$200 retail ($120 beta)

3 hours (beta)

About 80 multiple choice, plus hands-on labs

Google’s pricing pattern is a $99 foundational tier, a $125 associate exam and $200 for the professional exams I could confirm. Always check the specific exam page before paying, because prices and formats change and the other professional exams may differ.

How Exams Are Delivered

Google moved its testing partner from Kryterion to Pearson in early 2026. Kryterion’s last testing day was February 22, 2026, and the first Pearson exams ran from March 2, 2026. Old study notes and forum posts that mention Kryterion or Webassessor are out of date. Today you can sit an exam online with Pearson’s OnVUE remote proctoring or in a Pearson test center.

Booking runs through CM Connect, Google’s candidate portal. You choose an exam and delivery mode, then are signed in to Pearson automatically, with no separate Pearson account. Since October 29, 2026, CM Connect no longer accepts an email-and-password sign-in. You log in with Google Skills credentials, so link the two accounts before exam day.

A few policies worth knowing:

  • Cancellations and reschedules must happen at least 24 hours ahead for both test-center and online appointments. Miss that window and you forfeit the exam fee.

  • Results are pass or fail only. No numerical score is issued. You see a provisional result on screen, and Google confirms it after review, typically in 7–10 days.

  • If you fail, the Candidate Portal gives a section-level report showing the approximate share of the exam each section represented. You do not have to pass every section. Overall performance decides the result.

  • Corporate laptops are a bad idea for online proctoring. Pearson warns that firewalls and VPNs can interrupt the session.

Retake Rules

The retake rules differ by tier. Foundational exams allow up to ten attempts in a year, with at least 14 days between failures. Associate and Professional exams allow four attempts in two years, and the waits stretch with each failure:

Attempt

Wait after previous failure

2nd

14 days

3rd

60 days

4th

365 days

Every attempt counts, whatever the language or delivery method, and you pay each time. Registering under another name to get around the limits voids the result and can revoke certification.

Renewal and Recertification

Foundational and Associate certifications last three years. Professional certifications last two. Google emails reminders to the primary address on your CM Connect profile, so keep it current.

Renewing by exam. You can only retake an exam you currently hold inside the renewal window: 180 days before expiry for Foundational and Associate, 60 days before for Professional. Trying earlier gets the result rejected and can cost you your certification. You can still renew up to 30 days after your inactive date and keep your original Series ID. Later than that, you pay the full standard price and sit the standard exam, though your old Series ID can be restored afterward. When you first certify, Google issues a 50% renewal discount code, which you can find in CM Connect You can also prepare for the Amazon Web Services (AWS) certification questions.

Some certifications have a shorter renewal exam. Associate Cloud Engineer’s is one hour with 20 questions at $75. Professional Cloud Architect’s is one hour with 25 questions at $100, built around case studies. Google’s page also shows shorter renewal exams for Professional Data Engineer and Cloud Digital Leader. Once you pick a renewal route for those certifications you can’t switch after your first attempt, though Cloud Architect candidates can.

Renewing without an exam. A few certifications offer a continuing-education route. You complete designated learning activities in Google Skills during the final year of validity, and each renewal adds one year. Only some certifications qualify, so look for the Renew button on your Google Skills dashboard. You can also pass a beta exam while certified and have it count toward renewal.

How to Choose Your First Google Cloud Certification

Match the exam to your job, not to the exam’s reputation.

If you don’t build things for a living, start with Cloud Digital Leader (broad Google Cloud concepts) or Generative AI Leader (AI strategy and Google’s AI offerings). Both are the cheapest exams in the catalog and neither requires hands-on experience. Generative AI Leader is more current. Cloud Digital Leader is broader.

If you work with infrastructure or applications, Associate Cloud Engineer is the best first technical exam. It is mostly practical: setting up projects, IAM, deploying and monitoring workloads. A few months in the console and with the gcloud CLI matters more than anything you can read about it.

If you already have years of cloud experience, skip straight to a Professional exam that matches your role. Nothing forces the Associate step. In my experience, people who go to Cloud Architect directly from another cloud do fine if they spend time on Google-specific networking and IAM design, which differ from AWS and Azure in ways that surprise people.

If you are in data or ML, go directly to Professional Data Engineer or Professional Machine Learning Engineer. Data Practitioner is an Associate-level alternative if you want a gentler entry.

Practice tests help you find weak spots, but nothing replaces a few weeks of real hands-on time. Google’s free tier and $300 in new-customer credits make that cheap.

How Google Cloud Compares With AWS and Microsoft Azure

Most people picking a Google Cloud certification are also looking at AWS and Azure. This part is opinion, not an official comparison.

  • AWS has the largest certification ecosystem, with Foundational, Associate, Professional and Specialty tiers. If your employer runs mostly on AWS, an AWS credential usually has more local job-market pull. Google’s Cloud Architect is the rough counterpart to AWS Solutions Architect.

  • Microsoft Azure certifications (Fundamentals, Associate, Expert) tend to dominate organizations built around Microsoft 365 and Windows Server. Azure certificates are renewed annually through an online assessment, which is lighter than Google’s two- or three-year cycle.

  • Google Cloud is the smaller player overall, but its data, analytics, Kubernetes and AI credentials carry weight in companies that chose it for those strengths. It is also unusual in issuing only pass/fail results.

A sensible rule is to certify with whichever cloud your target employers use, then add Google Cloud if you are working on data and AI.

FAQs