Oracle Agentic AI Foundations Associate: Free Exam Guide

Oracle Agentic AI Foundations Associate is a free, entry-level certification from Oracle University for people who want to prove they understand AI agents and how they are built. Yes, it is really free: Oracle lists the exam price as Free on its official exam page. Also, the exam is 1Z0-1157-26, with 40 multiple choice questions in 60 minutes and a 65 percent pass mark. In particular, it suits developers, cloud engineers, data professionals, and students who want a first credential in agentic AI without paying for an exam.

Oracle Agentic AI Foundations: what the certification is

Oracle Agentic AI Foundations Associate (exam 1Z0-1157-26) is a free Oracle certification testing foundation knowledge of AI agents, LangChain, the Model Context Protocol, the OpenAI Agents SDK, OCI Enterprise AI Agents, and Oracle AI Database, in 40 questions over 60 minutes with a 65 percent pass mark.

An AI agent is software that can reason about a goal, pick tools, take actions, and adjust based on results, rather than simply answering one prompt. Indeed, that difference is the heart of this certification. In addition, the exam does not test one product alone. It covers the ideas behind agents, the frameworks used to build them, the protocol that connects them to tools, and the Oracle platform services that run them in production.

Oracle positions this as a foundations credential, which means it is a starting point, not an advanced architect exam. If you already build production agent systems every day, you will find much of it familiar. However, if you are new to agents, it gives you a structured path from basic concepts to a working agent, with a certificate at the end that you can show.

Who should take it

Take this exam if any of these describe you. You are a developer or cloud engineer who wants to add agentic AI to your skill set. You are a student or early-career professional who wants a free, recognized credential for your resume. Also, you work with Oracle Cloud Infrastructure and want to understand its enterprise agent services. Or you are simply curious about agents and want a guided, hands-on introduction with a clear finish line.

Oracle states no prior certification as a prerequisite. The recommended background is basic familiarity with large language model concepts, working knowledge of Python, and basic familiarity with Oracle Cloud Infrastructure concepts. However, treat those as recommendations, not entry requirements. In fact, if you can read basic Python and you know what a language model is, you have enough background to start.

Oracle Agentic AI Foundations: exam facts

Every fact in this table comes from Oracle University’s official exam page for 1Z0-1157-26. However, if any detail changes, the official page is the source that counts, so confirm the current facts there before you register.

Exam detail	What Oracle lists
Certification name	Oracle Agentic AI Foundations Associate
Exam number	1Z0-1157-26
Format	Multiple choice
Number of questions	40
Duration	60 minutes
Passing score	65 percent
Exam price	Free
Validated against	Oracle Cloud Infrastructure 2026
Credential earned	Agentic AI Certified Foundations Associate

Two pieces of exam math are worth knowing before you sit down. A 65 percent pass mark on 40 questions means you need at least 26 correct answers. Moreover, 60 minutes for 40 questions gives you 90 seconds per question on average. In fact, that pace is comfortable if you know the material, but it punishes getting stuck. Flag hard questions, move on, and return to them with your remaining time.

After the exam, Oracle shows your score on screen, sends an email with the pass or fail result, and records your full score report in Oracle CertView. In addition, confirm the current recertification terms on Oracle’s site at registration, since Oracle lists a Cloud Recertification policy for this exam but the details can change.

Oracle Agentic AI Foundations: domain weights and study time

Oracle publishes six exam domains with weights. For example, the weights tell you exactly where to spend your study time: half the exam sits in the two Oracle platform domains, so give them half your preparation.

Domain	Exam weight	Suggested study time
Introduction to AI Agents	15 percent	About 2 hours
LangChain for AI Agents	5 percent	About 45 minutes
Model Context Protocol (MCP) Fundamentals	15 percent	About 2 hours
OpenAI Responses API and Agents SDK	15 percent	About 2 hours
OCI Enterprise AI Agents	25 percent	About 3 hours
Agentic AI for Oracle AI Database	25 percent	About 3 hours

Here is what each domain actually covers. Introduction to AI Agents covers how agents differ from chatbots and rule-based workflows, the core components of an agent (language model, tools, and the orchestration loop), reasoning patterns such as Chain-of-Thought and ReAct, and safety considerations and guardrails. In addition, LangChain for AI Agents covers chat models, prompts, tools, and agents as building blocks, plus how reasoning and tool execution flow inside a LangChain agent. However, it carries only 5 percent, so learn it well but do not let it eat your week.

Model Context Protocol and the OpenAI Agents SDK

Model Context Protocol Fundamentals covers the role of MCP in standardizing how agents connect to external tools, its core components (hosts, clients, servers, tools, resources, and prompts), the JSON-RPC 2.0 message format, and transport options including stdio and Streamable HTTP. Furthermore, the OpenAI domain covers the Responses API, the Agents SDK primitives (Agent, Runner, Tool, Handoffs, and Guardrails), function calling, multi-agent design patterns, and how guardrails validate inputs, outputs, and agent actions.

The two heaviest domains

The two heaviest domains are where most candidates should spend their time. OCI Enterprise AI Agents covers the enterprise AI agent lifecycle, the OCI Enterprise AI Agents service, its building blocks (Responses API, tools, memory, and vector stores), deployment, and scaling. In addition, Agentic AI for Oracle AI Database covers Oracle AI Vector Search (the VECTOR data type, embeddings, and similarity search), the workflow from document chunking to retrieval, grounding agent responses in enterprise data, the no-code Private Agent Factory, Select AI for natural-language interaction with data, and the Autonomous AI Database MCP Server. If you study nothing else deeply, study these two domains deeply.

Oracle Agentic AI Foundations: free prep path step by step

The preparation is free, and Oracle’s recommended training is optional, not a prerequisite. First, follow these steps in order.

Step 1: Open the official exam page. Start at Oracle University’s exam page for 1Z0-1157-26 and read the objectives list. In fact, this is your syllabus. Every study session should map back to one of the six domains above.

Step 2: Work through Oracle’s free Agentic AI Foundations learning path. The path runs about six and a half hours of material and is built around doing, not just watching: you build an agent from scratch, use LangChain and the OpenAI Agents SDK, work with a real MCP server, and see where agents run in production on OCI Enterprise AI Agents and Oracle AI Database features. However, treat the hands-on parts as mandatory, even though the training itself is optional. Indeed, agent skills stick through building, not through reading about building.

Build, practice, and review

Step 3: Build one small agent yourself. Pick a simple, useful task, such as an agent that answers questions from your own notes or looks up information through a tool. As a result, you will meet the exact ideas the exam tests: the orchestration loop, tool selection, and guardrails. Indeed, one working agent teaches more than a second pass over the same videos.

Step 4: Take Oracle’s practice exam. Oracle offers a practice exam for this certification. Take it under timed conditions, 60 minutes, no notes. Next, review every wrong answer against the domain list and close those gaps before you book the real exam.

Step 5: Review the two 25 percent domains once more. Reread the OCI Enterprise AI Agents and Oracle AI Database material, and write down the key terms from memory: vector stores, embeddings, similarity search, chunking, Select AI, Private Agent Factory, and the MCP Server. Finally, if you can explain each one in a sentence, you are ready.

Oracle Agentic AI Foundations: 2-week study plan

This plan assumes about an hour a day, which fits around work or study. Alternatively, if you have more time, compress it into one week by doubling the daily sessions, but keep the order the same.

Days	Focus	What to do
Days 1 to 2	Introduction to AI Agents	Learn agent components, the orchestration loop, Chain-of-Thought, ReAct, and guardrails. Write a one-page summary from memory.
Day 3	LangChain for AI Agents	Cover chat models, prompts, tools, and agents. Build a tiny LangChain agent that calls one tool.
Days 4 to 5	Model Context Protocol Fundamentals	Learn hosts, clients, servers, tools, resources, and prompts, plus JSON-RPC 2.0 and both transport options. Connect an agent to a real MCP server.
Days 6 to 7	OpenAI Responses API and Agents SDK	Study Agent, Runner, Tool, Handoffs, and Guardrails. Build an agent with function calling and one handoff between two agents.
Days 8 to 10	OCI Enterprise AI Agents	Work through the OCI service: lifecycle, building blocks, memory, vector stores, deployment, and scaling. Build and run a basic agent on the service if you can access it.
Days 11 to 12	Agentic AI for Oracle AI Database	Study vector search end to end: chunking, embeddings, similarity search, and retrieval. Learn Select AI, Private Agent Factory, and the MCP Server.
Day 13	Practice exam	Take the practice exam timed at 60 minutes. Review every wrong answer against the domain list.
Day 14	Final review and registration	Revisit your weakest domain, reread your summaries, then register for the real exam through Oracle MyLearn.

Do not leave the Oracle domains to the last day

A common mistake is spending the first week on general AI news and agent theory and leaving the Oracle platform domains to the last day. Instead, reverse that instinct. However, the general ideas matter, but half the exam is OCI and Oracle AI Database, so those domains deserve the largest blocks in your calendar.

Oracle Agentic AI Foundations: how to register

Registration runs through Oracle MyLearn. The broad steps are: sign in with your Oracle account on the exam page for 1Z0-1157-26, enroll in the free learning path and practice exam if you have not already, then register for the exam itself, where the price shows as Free. In addition, confirm the exact screens and any identity or scheduling details on Oracle’s site at registration, since those steps can change and this guide does not claim details Oracle has not published on its exam page.

On exam day, keep the pace in mind: 90 seconds per question on average, 26 correct answers to reach the 65 percent pass mark. First, answer the questions you know first, flag the rest, and use your remaining time on the flagged set. Finally, your score appears on screen when you finish, and your full report lands in Oracle CertView.

Oracle Agentic AI Foundations: what it does and does not do

This certification does something specific and useful: it proves you understand agent fundamentals, the main frameworks and protocols, and Oracle’s enterprise agent platform at a foundation level, and it gives you a credential you can list. For example, for a student or early-career professional, that is a real resume signal, especially because it is free and therefore easy for employers to take at face value as self-driven learning.

What it does not do matters just as much. It does not guarantee a job, a salary increase, or a promotion. It does not make you an AI architect: that is a professional-level role with deeper design and production experience behind it. And it does not replace hands-on work. Moreover, employers will still ask what you built. The strongest use of this certificate is as proof that your foundation is solid, paired with one or two small agents you can show and explain.

Oracle Agentic AI Foundations vs Google Agentic Architect

Readers often ask how this compares with the Google Cloud Professional Agentic Architect path. The honest answer is that they sit at different levels. Oracle Agentic AI Foundations is a free, foundations-level exam: 40 multiple choice questions, 60 minutes, and a 65 percent pass mark, with free official preparation. Google’s path is a professional-level certification with a longer exam that includes hands-on labs and a paid registration, and its general availability registration was expected in mid-November 2026. For example, if you want a free first credential in agentic AI, start with Oracle. If you already work on Google Cloud and want a professional-level architect credential, Google is the next step. In addition, our full guide to the Google path is here: Google Cloud Agentic Architect Exam Guide.

Similarly, one related note for planners: NVIDIA’s NCP-AAI agentic AI certification has been announced, but registration is not open yet, so no booking guide for it exists here at this time.

If a free, structured starting point in agentic AI sounds right, this is one of the easiest yes decisions in certification: the exam costs nothing, the preparation costs nothing, and two focused weeks are enough to be ready. Finally, start with Oracle’s official exam page, work the free learning path in domain order, build one small agent, take the practice exam, and then register.