- GAF here means Generative AI Foundations, a Pearson exam delivered through Certiport within Critical Career Skills.
- The exam runs 50 minutes and requires a scaled score of 700 on Certiport's 1-1000 scale.
- Four unweighted domains: methods, basic prompt engineering, prompt refinement, and ethics, law, and societal impact.
- Certiport vouchers work only at Certiport Authorized Testing Centers; Mindhub vouchers follow separate country rules.
What GAF Actually Is
The acronym GAF gets used by several unrelated credentials, so it is worth being precise from the first sentence. On this site, GAF means Generative AI Foundations: a proctored, computer-based knowledge assessment that tests whether a candidate understands how generative AI works and can use it responsibly and effectively. It is not a coding exam, not a vendor product certification, and not tied to any single AI tool.
The exam is described as brand-agnostic. That matters for preparation, because you are being tested on concepts and practices that hold across chatbots, image generators, and other generative systems rather than on the menus of one particular product. If you want the shorter definitional versions of this topic, our pages on what GAF certification is and what GAF stands for cover the naming question directly.
Who Issues It and Where It Sits
Generative AI Foundations is issued by Pearson. It is developed and administered through Certiport, a Pearson VUE business, and it belongs to the Critical Career Skills (CCS) family of certifications. Pearson announced the credential on October 1, 2024, and the objectives document that defines its content carries a 2024 copyright. That copyright identifies the source document; it should not be read as a formal exam-version label.
Because Certiport also runs many classroom-oriented certification programs, the exam is commonly encountered through schools, colleges, and workforce programs. Individual candidates can also take it, subject to the delivery and voucher rules covered below.
The Four Exam Domains
Certiport's exam objectives organize the content into four domains. The outline does not publish percentage weights, so any site claiming precise weightings for these domains is guessing. What you can rely on is the structure itself, and it follows a sensible arc: understand the technology, learn to prompt it, learn to improve your prompts, and understand the responsibilities that come with using it. Our complete guide to all four content areas goes deeper on each; here is the orientation.
Domain 1: Generative AI Methods and Methodologies
The conceptual foundation. Candidates need to explain what generative AI is, how it differs from other approaches to AI, and how generative systems produce text, images, and other content.
- Core vocabulary: models, training data, inputs, outputs, and generation
- What generative systems do well and where they predictably fail
- The kinds of content generative tools can produce and common use cases
Domain 2: Basic Prompt Engineering
The practical entry point. This domain tests whether you can write prompts that get usable results from a generative tool.
- Giving clear instructions, relevant context, and a defined task
- Specifying audience, tone, format, and constraints
- Recognizing why vague prompts produce vague or off-target output
Domain 3: Prompt Refinement
Where first-pass prompting becomes a skill. Candidates should be able to look at an unsatisfying response and decide what to change.
- Iterating on a prompt based on what the output got wrong
- Adding constraints, examples, or clarifications to steer results
- Evaluating output quality and deciding when a response is good enough to use
Domain 4: Ethics, Law, and Societal Impact
The responsible-use domain. Expect scenario thinking rather than pure definitions.
- Bias, fairness, and the reliability limits of generated content
- Intellectual property, privacy, and legal considerations when using generative tools
- Broader effects on work, education, and society, and what responsible use looks like
Notice the proportions of the outline: two of the four domains are about prompting. Candidates who treat the exam as "an AI theory test" tend to underprepare for the hands-on judgment that Domains 2 and 3 reward.
Exam Format, Timing, and Scoring
The exam is a proctored, computer-based knowledge assessment. You select answers, can flag questions, and can review your responses before final submission. The timed portion is 50 minutes, and that figure excludes the tutorial and administrative seat time, so your total time in the chair will be longer than 50 minutes.
Two details deserve honesty rather than false precision:
- Question count: Certiport's 2024 GAF tutorial specifies 40-45 questions, while the current Mindhub voucher page estimates 45-50. Certiport's scoring policy says totals differ by exam form. Plan for a range, not a fixed number.
- Scoring: The passing score is 700 on Certiport's 1-1000 scaled-score system. This is not a percentage of questions answered correctly, so you cannot convert it into "I need to get X questions right." Our page on the GAF passing score explains how to think about it.
| Exam Detail | What the Issuer Materials Say |
|---|---|
| Timed duration | 50 minutes, excluding tutorial and administrative time |
| Question count | 40-45 per Certiport tutorial; 45-50 estimated on Mindhub; varies by form |
| Passing score | 700 on a 1-1000 scaled score |
| Format | Proctored, computer-based; answer selection, flagging, review before submission |
| Domain weights | Not published in the objectives outline |
If you are wondering how demanding that adds up to be, see our difficulty guide. For published pass-rate claims, check what the data does and does not show in our GAF pass rate analysis before trusting any number you find.
Registering and Paying
Registration is where candidates most often get tripped up, because there are two voucher routes and they are not interchangeable.
Certiport vouchers
In the U.S., Certiport lists a single-exam voucher at USD 72 and a voucher bundled with a retake at USD 83. A Certiport Authorized Testing Center (CATC) may add its own proctoring charge on top. These vouchers are valid only for CATC delivery, whether in person or remote, and cannot be redeemed through OnVUE or a Pearson VUE Testing Center.
Mindhub vouchers
Pearson's separate Mindhub voucher is listed at USD 85, and Mindhub Pro lists both web and member prices as USD 85. Mindhub vouchers carry country restrictions and must match the country selected at checkout.
| Option | Listed Price (USD) | Key Condition |
|---|---|---|
| Certiport single-exam voucher | 72 | CATC delivery only; possible proctoring fee |
| Certiport voucher with retake | 83 | CATC delivery only; possible proctoring fee |
| Mindhub voucher | 85 | Country must match checkout selection |
Candidates under 18 need parental consent under CATC rules, and valid identification is required. Books and personal devices must remain inaccessible during the exam. For the full fee picture, see our GAF certification cost breakdown, and for scheduling logistics, our page on GAF exam dates and scheduling.
Who the Exam Is Built For
The exam is described as best suited for candidates ages 14 and older, and no bachelor's degree is required. That puts it squarely in the entry-level, broad-access category: students, career changers, office professionals, and anyone who needs documented evidence that they understand generative AI.
The issuer's stated preparation expectation is 150 combined hours of instruction and hands-on use of generative AI tools, plus familiarity with productivity applications. Two clarifications keep that figure from being misread. First, it describes how much preparation the issuer expects, not how long the exam takes. Second, it is not a mandatory course length; you do not need to enroll in a specific 150-hour class. The practical takeaway is that hands-on tool use counts, so real practice with generative AI tools is not optional busywork. It is the point.
For a fuller treatment of eligibility, see GAF requirements.
How Long the Credential Lasts
CCS certifications earned on or after June 18, 2025 are valid for five years. Recertification requires purchasing a new voucher and passing the current exam, so the renewal path is simply re-testing. Earlier nonexpiring certifications retain their existing status.
Practically, this means a GAF credential is not a one-and-done line item. Given how quickly generative AI tools change, a five-year window with re-testing against the then-current exam is a reasonable design, and it is worth factoring into any long-term value calculation.
Where the Credential Fits in a Career
It helps to be realistic about what a foundations-level credential does. GAF signals baseline literacy: you understand what generative AI is, you can prompt it competently, you can improve imperfect output, and you know the ethical and legal landmines. That is valuable in roles where generative AI is a tool rather than the product, such as administrative, marketing, content, education, customer support, and general knowledge-work positions.
It is not a substitute for machine learning engineering skills, and it does not by itself qualify someone for specialist AI development roles. Treat it as a credible entry point that complements existing experience. We deliberately avoid quoting salary figures here because we will not invent numbers; for a measured discussion of earnings context, read our GAF salary guide, browse GAF jobs, or weigh the tradeoffs in is the GAF certification worth it.
Key Takeaway
Position GAF as proof of applied literacy. On a resume, pair it with a concrete example of how you used generative AI at work or in a project; the credential plus a demonstrable result is far stronger than the credential alone.
A Domain-Ordered Preparation Plan
Because the domains build on each other, order matters more than intensity. Here is a four-week sequence that follows the exam's own logic. Adjust the pace to your starting point and to the hands-on hours you have already logged. For a fuller plan, use our GAF study guide.
Domain 1: Methods and Methodologies
- Build your vocabulary first; every later domain assumes it
- Experiment with a text tool and an image tool to see generation behavior firsthand
- Note where outputs are confidently wrong, since this feeds Domain 4
Domain 2: Basic Prompt Engineering
- Write the same request five ways and compare results
- Practice specifying audience, format, tone, and constraints
Domain 3: Prompt Refinement
- Take weak outputs and diagnose what to change in the prompt
- Practice judging whether a response is usable as-is
Domain 4 plus full review
- Work through bias, privacy, and intellectual property scenarios
- Take timed practice sets and review every miss by domain
For quick end-of-prep review, our GAF cheat sheet condenses the must-know facts, and you can run realistic timed drills on the GAF practice test site. Practice questions are most useful when you review why each wrong option is wrong, since the exam's scenario items reward judgment over memorization. If you want a structured question bank to time yourself against the 50-minute limit, start from the main practice test.
Frequently Asked Questions
No. On this site, GAF means Generative AI Foundations, issued by Pearson and administered through Certiport. Other credentials share the abbreviation but have different issuers, fees, and content, so do not mix their details with this exam.
The timed portion is 50 minutes, excluding the tutorial and administrative time. You need a scaled score of 700 on Certiport's 1-1000 scale. That is a scaled score, not a percentage of questions answered correctly.
Sources differ. Certiport's 2024 tutorial specifies 40-45 questions, while the Mindhub voucher page estimates 45-50. Certiport's scoring policy states totals vary by exam form, so prepare for a range rather than a fixed count.
No. Certiport vouchers are valid only for delivery at a Certiport Authorized Testing Center, in person or remotely. They cannot be redeemed through OnVUE or a Pearson VUE Testing Center, so confirm compatibility before purchasing.
CCS certifications earned on or after June 18, 2025 are valid for five years. Recertifying means buying a new voucher and passing the current exam. Earlier nonexpiring certifications keep their existing status.