1Z0-1158-26 study guide · 2026

The 6 1Z0-1158-26 domains, explained

6 areas and 30 topics. Oracle does not publish weights for these areas, so each gets equal space here. Each guide explains what the area covers, how it is assessed and where candidates slip.

Start with Domain 1 See all domains
Questions
50
Time
90 minutes
Passing score
68%
Outline
1Z0-1158-26 exam topics (2026 release; section names from secondary source, unverified)

Exam blueprint

Oracle does not publish weights, so every area is drawn the same size.

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  1. 1 Applying Large Language Model Concepts
  2. 2 Applying Agentic AI Concepts
  3. 3 Configuring OCI Enterprise AI Models
  4. 4 Building OCI Enterprise AI Agents
  5. 5 Building and Deploying Hosted Agents
  6. 6 Governing and Optimizing OCI Enterprise AI

All 6 domains

30 official topics 64–86 hrs suggested study

  1. Domain 01

    Applying Large Language Model Concepts

    The language model groundwork for 1Z0-1158-26: how decoder models turn tokens into text, which inference parameters shape output, how prompts and retrieval steer answers, and when customization is worth it.

    Inside this domain

    • Explain how large language models generate text
    • Control output with inference parameters
    • Apply prompt engineering techniques
    • Use embeddings and retrieval-augmented generation
    • and 1 more topics

    Watch for Temperature fixes accuracy

    Topics
    5
    Questions est.
    ~5
    Study time est.
    8–10 hrs
  2. Domain 02

    Applying Agentic AI Concepts

    Vendor-neutral agent design for 1Z0-1158-26: how an agent plans and acts in a loop, how tools and MCP servers extend it, how memory carries context, and when to add workflows or human review.

    Inside this domain

    • Explain the agent loop and autonomy levels
    • Describe tools and function calling
    • Explain the Model Context Protocol
    • Manage agent memory and context
    • and 1 more topics

    Watch for Model runs the tool

    Topics
    5
    Questions est.
    ~10
    Study time est.
    10–14 hrs
  3. Domain 03

    Configuring OCI Enterprise AI Models

    How 1Z0-1158-26 expects you to choose and configure models on OCI: on-demand access, OpenAI-compatible calls, dedicated AI clusters for hosting and training, custom fine-tuned models and imported open-weight models.

    Inside this domain

    • Select OCI Enterprise AI models and access modes
    • Call models through OpenAI-compatible endpoints
    • Provision dedicated AI clusters
    • Fine-tune base models into custom models
    • and 1 more topics

    Watch for Custom model is live after training

    Topics
    5
    Questions est.
    ~10
    Study time est.
    12–16 hrs
  4. Domain 04

    Building OCI Enterprise AI Agents

    The largest hands-on block of 1Z0-1158-26: building agents with the OpenAI-compatible OCI Responses API, its four tool types, files and vector stores, conversation state and memory, and natural-language access to enterprise data.

    Inside this domain

    • Build agent workflows with the OCI Responses API
    • Use built-in tools in the Responses API
    • Ground agents with Files and Vector Stores
    • Manage conversation state and agent memory
    • and 1 more topics

    Watch for Any tool type works

    Topics
    5
    Questions est.
    ~13
    Study time est.
    16–20 hrs
  5. Domain 05

    Building and Deploying Hosted Agents

    Running your own agent runtime on OCI-managed infrastructure for 1Z0-1158-26: packaging it as a container, defining a hosted application and its deployments, choosing endpoints and authentication, and using sandboxes.

    Inside this domain

    • Package agent runtimes as container images
    • Configure hosted applications
    • Create and manage deployments
    • Configure authentication for hosted agents
    • and 1 more topics

    Watch for Push updates the agent

    Topics
    5
    Questions est.
    ~7
    Study time est.
    10–14 hrs
  6. Domain 06

    Governing and Optimizing OCI Enterprise AI

    Keeping OCI Enterprise AI workloads safe, visible and affordable for 1Z0-1158-26: IAM and API key choices, guardrails for content, prompt injection and PII, audit and monitoring, data handling and cost tuning.

    Inside this domain

    • Secure Enterprise AI resources with IAM
    • Apply guardrails to model inputs and outputs
    • Monitor, audit and observe agent workloads
    • Optimize cost and performance
    • and 1 more topics

    Watch for Inform mode protects users

    Topics
    5
    Questions est.
    ~5
    Study time est.
    8–12 hrs

Suggested study order

No official weights are published, so follow the outline order, across about 64–86 hrs of study.

  1. 1 Applying Large Language Model Concepts 8–10 hrs
  2. 2 Applying Agentic AI Concepts 10–14 hrs
  3. 3 Configuring OCI Enterprise AI Models 12–16 hrs
  4. 4 Building OCI Enterprise AI Agents 16–20 hrs
  5. 5 Building and Deploying Hosted Agents 10–14 hrs
  6. 6 Governing and Optimizing OCI Enterprise AI 8–12 hrs

Try one

Practice questionDomain 1: Applying Large Language Model Concepts

A claims team uses an LLM to label emails into five categories. Labels vary between identical runs and the model sometimes invents a sixth category. Which change should they try first?

One trap from every domain

The answer that looks right, and the clue that gives it away.

  • Domain 1

    Temperature fixes accuracy

    Clue The scenario says facts are absent from training data or recently changed.

  • Domain 2

    Model runs the tool

    Clue The question asks which component needs database credentials.

  • Domain 3

    Custom model is live after training

    Clue The scenario reports errors calling a newly trained model.

  • Domain 4

    Any tool type works

    Clue Options list tool names not in the four supported types.

  • Domain 5

    Push updates the agent

    Clue The scenario asks how to release or roll back code.

  • Domain 6

    Inform mode protects users

    Clue Requirement words such as must reject or must prevent.

Other Artificial Intelligence & Machine Learning certifications

Study guides for exams that overlap with 1Z0-1158-26.

Practise what you read

1Z0-1158-26 domains: common questions

How many domains are in the 1Z0-1158-26 exam?

1Z0-1158-26 has 6 domains with 30 official topics between them: Applying Large Language Model Concepts, Applying Agentic AI Concepts, Configuring OCI Enterprise AI Models, Building OCI Enterprise AI Agents, Building and Deploying Hosted Agents and Governing and Optimizing OCI Enterprise AI.

What is the 1Z0-1158-26 exam format?

Questions: 50. Time allowed: 90 minutes. Passing score: 68%.

How long should I study for 1Z0-1158-26?

Our planning estimate is 64–86 hrs across all domains, split by weight. It is an editorial estimate, not an official figure, so adjust it to your experience and to how you score on practice questions.

Where should I start?

Work through the areas in order, since each builds on the one before. Use the checklist on each page to confirm you can do what it asks before moving on.

Which outline is this guide based on?

This guide follows 1Z0-1158-26 exam topics (2026 release; section names from secondary source, unverified), published by Oracle. It was last reviewed in October 2026.

Domain names, weights and topics come from Oracle's official outline (1Z0-1158-26 exam topics (2026 release; section names from secondary source, unverified)). Question counts and study hours are editorial planning estimates, not official figures. Last reviewed October 2026.