1.1 Current Exam Facts

Microsoft Certified: AI Agent Builder Associate AB-620

Track your progress as you read

Finished sections get a check mark.

1 Introduction & Exam Overview

1.1 Current Exam Facts 1.2 Candidate Profile & Prerequisites 1.3 Skills Map & Study Strategy

2 Plan Integration, Identity & Channels

2.1 Plan Integration with Enterprise Systems 2.2 Plan Identity Strategy 2.3 Plan Channels and Deployment

3 Responsible AI, Security & Reuse

3.1 Plan Responsible AI Strategy 3.2 Security and Governance Considerations 3.3 Plan Reusable Agent Components 3.4 Design Agents for Internal or External Audiences

4 Agent Flows Fundamentals

4.1 Create an Agent Flow 4.2 Human-in-the-Loop Agent Flows 4.3 Configure Actions and Connectors in Flows

5 Agent Flow Monitoring & Resilience

5.1 Monitor Agent Flows 5.2 Input and Output Parameters 5.3 Error Handling in Agent Flows

6 Topics Configuration Core

6.1 Add Agent Flows to a Topic 6.2 Configure Agent Response Formatting 6.3 Add Tools to a Topic

7 Advanced Topic Responses

7.1 Custom Prompts for Advanced Responses 7.2 Custom Knowledge Sources for Advanced Responses 7.3 API and Send HTTP Request Responses

8 Generative Answers, Cards & Variables

8.1 Configure the Generative Answers Node 8.2 Configure Adaptive Cards 8.3 Manage Variables

9 Enterprise Knowledge Sources

9.1 Connect to Copilot Connectors 9.2 Connect to Power Platform Connectors 9.3 Connect to Azure AI Search

10 Add Tools to Agents

10.1 Configure and Monitor Computer Use 10.2 Configure MCP Tools 10.3 Tools via Existing Custom Connectors 10.4 Add REST APIs to an Agent

11 Multi-Agent Collaboration

11.1 Design Multi-Agent Solutions 11.2 Integrate a Foundry Agent 11.3 Integrate an Existing Copilot Studio Agent

12 Fabric Data Agents & A2A

12.1 Integrate a Fabric Data Agent 12.2 Multi-Agent Solutions with A2A Protocol

13 Integrate Agents with Azure

13.1 Generative Answers with Azure AI Search & Foundry 13.2 Foundry Model Catalog for Custom Prompts 13.3 Monitor Agents with Application Insights

14 Evaluate Agent Performance

14.1 Create a Test Set 14.2 Choose an Evaluation Method 14.3 Review Test Results

15 Application Lifecycle Management

15.1 Create a Solution 15.2 Add Existing Agents to a Solution 15.3 Create and Use Environment Variables 15.4 Power Platform Pipelines for Agents

1.1 Current Exam Facts

Key Takeaways Explain This Topic AI

Exam AB-620 earns Microsoft Certified: AI Agent Builder Associate; Microsoft Learn titles the study guide Designing and Building Integrated AI Solutions in Copilot Studio.

Passing requires a scaled score of 700 or greater on a 1000-point scale; exam duration is 120 minutes (plus 30 minutes if your preferred language is unavailable).

Microsoft does not publish a fixed question count for AB-620; register through Pearson VUE at a test center or online proctoring.

Skills measured weights: Plan and configure agent solutions 30–35%, Integrate and extend agents in Copilot Studio 40–45%, Test and manage agents 20–25%.

Standard associate pricing is region-based (about USD 165 in many markets); beta attempts may delay score reporting until scoring is finalized.

Last updated: August 2026

Why exam facts matter before you open Copilot Studio

Exam AB-620 is the single required assessment for Microsoft Certified: AI Agent Builder Associate . Microsoft Learn titles the official study guide Designing and Building Integrated AI Solutions in Copilot Studio , and the same page's description renders it as Designing and Building Integrated AI Agent Solutions in Copilot Studio — treat both as the same exam rather than memorizing one string. If you skip the logistics—scaled scoring, domain weights, Pearson VUE delivery, and what Microsoft does not publish—you will misallocate study hours and walk into scenario stems unprepared for how the exam actually measures agent design work.

This section is the factual baseline for the rest of the guide. Every percentage, time limit, and audience statement below comes from Microsoft Learn study-guide and certification pages. Where Microsoft leaves a detail unpublished (notably a fixed item count), this guide does not invent a substitute number.

Credential and product surface

At a glance, Microsoft classifies the credential as:

Attribute Official positioning
Credential Microsoft Certified: AI Agent Builder Associate
Exam code AB-620
Level Intermediate (associate)
Primary product Microsoft Copilot Studio
Related products Microsoft Power Platform, Microsoft 365 Copilot, Azure / Microsoft Foundry
Roles Developer, app maker / advanced builder
Subjects Generative AI, solution design

You are not certifying as a pure Azure AI engineer or a Power Platform functional consultant alone. You are certifying that you can plan, configure, integrate, extend, test, and manage enterprise-grade agents that live primarily in Microsoft Copilot Studio and connect outward to enterprise systems, Foundry models, Fabric data agents, connectors, REST APIs, and multi-agent protocols.

Format, scoring, and time

Detail What Microsoft publishes
Passing score 700 or greater on a 1000 -point scaled score
Duration 120 minutes of exam time
Language accommodation If the exam is not available in your preferred language , you may request +30 minutes
Question count Not published as a fixed number—do not memorize third-party guesses as official
Delivery Pearson VUE (test center or online proctoring, subject to Microsoft/Pearson policies)
Proctoring Proctored assessment; interactive components may appear

Price Region-based ; standard associate exams are commonly about USD 165 in many markets—confirm at scheduling time

Exam trap: A scaled score of 700 is not “70% of items correct.” Microsoft uses psychometric scaling. Study for mastery of skills measured, not for a raw-count target that does not exist on the official pages.

During beta , score reporting can be delayed until Microsoft finalizes scoring. If you sit a beta form (AB-620 beta activity around April 2026 ), plan career or project milestones around possible lag—not next-day results. After general availability, standard score reporting policies apply.

Skills measured weights (your study budget)

Microsoft publishes three domains with percentage ranges:

Plan and configure agent solutions — 30–35%

Solution planning (enterprise integration, identity, channels, responsible AI, security/governance, reusable components, internal vs external audiences), agent flows (including human-in-the-loop), and advanced topic configuration (tools, prompts, knowledge, HTTP, generative answers, adaptive cards, variables).

Integrate and extend agents in Copilot Studio — 40–45%

The largest domain. Enterprise knowledge (Copilot connectors, Power Platform connectors, Azure AI Search), tools (computer use, MCP , custom connectors, REST APIs), multi-agent collaboration (Foundry agents, existing Copilot Studio agents, Fabric data agents, A2A ), and Azure integration (Foundry model catalog prompts, Application Insights monitoring).

Test and manage agents — 20–25%

Evaluation (test sets, evaluation methods, result review) and ALM (solutions, environment variables, Power Platform Pipelines).

Treat integrate and extend as the default priority for deep lab time. A candidate who only masters topic authoring without MCP tools, multi-agent design, or enterprise knowledge connections is under-preparing for nearly half the exam.

Languages, retakes, and registration hygiene

Microsoft lists AB-620 in multiple languages, including English, Arabic (Saudi Arabia), Chinese (Simplified and Traditional), French, German, Indonesian, Italian, Japanese, Korean, Portuguese (Brazil), Russian, and Spanish. Availability can lag English for updates; that is when the +30 minute language request becomes relevant.

Retake policy (Microsoft Credentials standard): after a first failure , you may retake after 24 hours . Subsequent retakes follow longer wait rules—read the current retake policy before burning voucher money on a second attempt the same week.

Microsoft strongly recommends registering with a personal Microsoft account (MSA) . If you register only under a work/school Azure AD account and later leave that organization, exam records can become unrecoverable . Link your certification profile to Microsoft Learn so scheduling, score history, and annual renewal stay under your control.

What “pass” means for the credential

Passing AB-620 awards Microsoft Certified: AI Agent Builder Associate . Like other Microsoft associate credentials, the certification expires annually and can be renewed free by completing the online renewal assessment on Microsoft Learn when it becomes available for this credential. Passing does not replace tenant admin rights, security review, or responsible-AI governance in a customer environment—it validates your ability to design and build integrated agent solutions under exam conditions.

Official sources to bookmark
AB-620 study guide — skills measured, audience profile, resource links
AI Agent Builder Associate certification page — overview, scheduling, duration, assessed skills
Microsoft exam sandbox — practice the UI and question interaction styles before exam day
Exam scoring and score reports documentation — confirmation of the 700 pass threshold
Realistic exam-day scenario

You schedule AB-620 for 120 minutes with English as the delivery language. On screen you face a multi-part case: an enterprise wants a Copilot Studio agent that answers from ServiceNow knowledge, calls a custom connector for order status, and hands off specialized analytics to a Fabric data agent. Domain weights tell you why this stem is fair game— integration and extension is the heavyweight domain, while planning identity and responsible AI still sits in domain 1 and ALM packaging sits in domain 3. Your job is not to invent a chat demo; it is to choose the architecture, tools, and governance options that match Microsoft’s published skills outline.

Quick success checklist for this section

Memorize 700/1000 , 120 minutes , Pearson VUE , and the three domain weight bands .

Refuse to “study to a fake question count.”

Budget the majority of deep practice for integrate and extend (40–45%) .

If you sit beta, expect delayed scores and plan accordingly.