Skip to main content
  1. Public practice questions
  2. AB-620 Practice Exams | AI Agent Builder Associate
  3. AB-620 Practice Exam 1

Question 15 — AB-620 Practice Exam 1

For each managed Prompt builder model description, drag the appropriate model category into the Correct Match column.

Each item may be used once, more than once, or not at all.

Option Correct Match
Includes GPT-4.1 mini, is the default choice, suits moderately complex tasks where speed matters, and is billed at the basic rate [Gap 1]
Includes GPT-4.1, supports more complex or multimodal tasks such as image and document analysis, and is billed at the standard rate [Gap 2]
Includes GPT-5 reasoning, targets reasoning-intensive work such as nuanced decisions and sophisticated problem solving, and is billed at the premium rate [Gap 3]

  1. General
  2. Generative answers
  3. Custom data
  4. Mini
  5. Deep

Correct answers

  1. Gap 1: Mini
  2. Gap 2: General
  3. Gap 3: Deep

Explanations

Correct Matches

Mini – This category includes GPT-4.1 mini, the default Prompt builder model. It is intended for moderately complex tasks where efficiency and speed are important and uses the basic Copilot Credits rate.

General – This category includes GPT-4.1 and provides stronger capability for complex or multimodal work, including image and document analysis and advanced content creation. It uses the standard rate.

Deep – This category includes GPT-5 reasoning and targets reasoning-intensive tasks such as nuanced decision-making, data analysis, and sophisticated problem-solving. It uses the premium rate.

Unused Item(s)

Custom data – Custom data is a knowledge-input mechanism for generative answers nodes rather than a managed model category in Prompt builder.

Generative answers – Generative answers is a knowledge-retrieval capability rather than a Prompt builder model category.