OCI AI Foundations
Oracle Cloud Infrastructure AI Foundations
More study resources
Cheat sheet
Keep the key ideas within reach.
ExploreKey terms
Get comfortable with the vocabulary.
ExploreStudy plan
Give each week a clear focus.
ExploreCommon mistakes
Understand the traps and how to avoid them.
ExploreCase studies
Put your knowledge to work in context.
ExploreExam-day guide
Prepare for the day ahead.
ExploreAbout this exam
OCI AI Foundations is an entry-level certification from Oracle that validates whether you can talk sensibly about artificial intelligence and machine learning in an Oracle Cloud Infrastructure context. It is aimed at people who are new to AI rather than specialists: developers and data analysts moving into AI work, solution architects and sales or pre-sales engineers who need to position OCI AI services, and business-facing staff who have to reason about what a model can and cannot do. The exam is conceptual first and product second. Expect questions on core ML vocabulary and the supervised/unsupervised distinction, how neural networks and deep learning fit together, how large language models are built and steered with prompt engineering and retrieval augmentation, and - heavily - when to reach for each OCI AI service: Language, Vision, Speech, Document Understanding, Anomaly Detection and Forecasting. A smaller share covers data science tooling and AI infrastructure on OCI, plus responsible AI concerns such as bias, explainability and data privacy. Preparation works best in layers: read the official OCI AI Foundations learning path and service documentation, then open a tenancy and call a few services so the product names stop being abstract. This page is independent study material and is not affiliated with or endorsed by Oracle.
What you'll study
About OCI AI Foundations preparation
What is the exam format?
It is a multiple-choice certification exam delivered through Oracle's testing partner, taken either at a test centre or online with a proctor. Oracle publishes the current delivery options, item types and accommodations on its official certification page, which is the only place to confirm details before you book.
How many questions are there and how long do I get?
Oracle does not fix these numbers in public study material, and they can change between exam versions. Rather than rely on second-hand figures, check the current exam page on Oracle's certification site, which states the live question count and time allowance for this exam.
How much does it cost and what score do I need to pass?
Both the fee and the passing score are set by Oracle and revised periodically, so any number quoted in a blog or forum should be treated as out of date. The official Oracle certification page lists the current price and the scaled passing score for this exam.
Do I need to know how to code?
No. The exam tests conceptual understanding and service selection, not implementation. That said, candidates who have written even a few lines against an SDK or tried a notebook tend to answer scenario questions faster, because they recognise what a service returns and where its limits lie.
Do I need hands-on OCI experience?
It helps more than any flashcard deck. A free tier tenancy is enough to try Language, Vision and Speech, and to see what a model endpoint or a Data Science notebook actually looks like. Half a day of clicking through the console makes service-comparison questions much easier to reason about.
In what order should I study the topics?
Start with general AI and ML vocabulary, since every later domain borrows it. Then cover neural networks and deep learning, move to generative AI and prompt engineering, and only then study the individual OCI AI services. Finish with data science infrastructure and responsible AI, which are lighter and easier to absorb at the end.
What trips most candidates up?
Two things. First, memorising service names without knowing which one fits a scenario - the exam asks you to choose, not define. Second, blurring generative AI techniques: retrieval augmentation, fine-tuning and prompt design solve different problems and are frequently confused under exam pressure.
Is it a prerequisite for other OCI certifications?
It is a standalone foundational credential rather than a formal gate. Oracle offers separate, more advanced certifications in data science, AI and cloud architecture, and this exam is a sensible warm-up for those, but you should confirm current prerequisites on Oracle's certification site rather than assume a chain.
How hard is it, and what if I fail?
Most candidates with a week or two of structured study and some console time find it manageable, because the questions are conceptual rather than mathematical. If you do not pass, Oracle sets the retake waiting period and any fees; check the official certification policy page for the current rules.