AWS Generative AI Developer – Professional Exam Guide (AIP-C01): Domains & Format

AIP-C01 Exam Format at a Glance

The AWS Certified Generative AI Developer – Professional (AIP-C01) exam validates your ability to integrate foundation models into production applications on AWS. Here are the official details from the AWS exam guide:

For the full preparation roadmap and study timeline, pair this with our AWS Certified Generative AI Developer – Professional study guide.

Question Types

AIP-C01 uses two question formats:

• Multiple choice — one correct response and three distractors.
• Multiple response — two or more correct responses out of five or more options. You must select all correct responses to get credit (no partial credit).

There are 65 scored questions that determine your result, plus 10 unscored questions AWS uses to trial future content. The unscored questions are not flagged, so treat every question as if it counts. Unanswered questions are scored as incorrect, and there is no penalty for guessing — so never leave a blank.

The Five Content Domains

The exam is organized into five weighted domains. Here's what each one tests.

Domain 1: Foundation Model Integration, Data Management, and Compliance — 31%

The single largest domain. It covers:

• Selecting foundation models from Amazon Bedrock (Anthropic Claude, Meta Llama, Amazon Titan, Cohere) based on capability, context window, latency, and cost
• Designing solutions with vector stores, Retrieval Augmented Generation (RAG), and knowledge bases
• Integrating FMs into applications and business workflows
• Data management and compliance considerations for GenAI data pipelines

This is where most exams are won or lost. Know Bedrock's model catalog and how to architect a RAG solution cold.

Domain 2: Implementation and Integration — 26%

The second-heaviest domain. It covers:

• Implementing model inference with the Bedrock API (InvokeModel, Converse) and SageMaker JumpStart
• Prompt engineering and management — system prompts, few-shot learning, chain-of-thought, output parsing
• Building agentic AI solutions with Bedrock Agents, Lambda for tool use, and Step Functions for orchestration

Together, Domains 1 and 2 make up 57% of the exam — prioritize them.

Domain 3: AI Safety, Security, and Governance — 20%

A substantial domain that's easy to underestimate:

• Security and identity: IAM policies for Bedrock, VPC endpoints, encryption
• Responsible AI: content filtering, toxicity detection, Bedrock Guardrails, PII detection with Comprehend
• Governance for production GenAI workloads

Domain 4: Operational Efficiency and Optimization for GenAI Applications — 12%

Smaller, but high-yield because the topics are concrete:

• Cost optimization: caching, model selection by cost-performance, provisioned throughput vs. on-demand
• Optimizing for cost, performance, and business value
• Monitoring with CloudWatch metrics and logs

Domain 5: Testing, Validation, and Troubleshooting — 11%

The smallest domain, covering the lifecycle's tail end:

• Evaluating foundation models for quality and responsibility (automated metrics and human review)
• Troubleshooting inference issues, latency, and token-limit errors
• Error-handling patterns: retries, fallbacks, graceful degradation

What's Out of Scope

The exam guide explicitly lists tasks that are not tested, which helps you avoid over-studying:

• Model development and training
• Advanced ML techniques
• Data engineering and feature engineering (beyond RAG needs)

AIP-C01 is a developer exam about using foundation models, not a data-science exam about building them.

How to Use This Guide

1. Confirm you meet the recommended prerequisites before committing.
2. Understand how the 750 passing score is calculated so you can set practice-exam targets.
3. Budget for the exam cost and any retakes.
4. Follow the full study timeline and resource list in the AWS Certified Generative AI Developer – Professional study guide.