Validate your ability to build production generative AI applications on AWS. This exam covers foundation model integration, RAG architectures, prompt engineering, agentic AI, vector stores, AI safety and governance, cost optimization, and testing and troubleshooting of GenAI solutions.
No sign-up — instant feedback after every question
Get instant feedback after each question. See explanations and track your progress as you learn.
Timed session with no feedback until the end. Experience real exam conditions and test your knowledge.
10 questions · pick a topic · 5 minutes · shareable
Every Bedrock feature, RAG and agent pattern, and safety / observability decision the AIP-C01 exam actually tests — organised by the five exam domains with reference architectures and runbooks
The AWS Certified Generative AI Developer - Professional (AIP-C01) is AWS's newest professional-level certification, validating your ability to build production-grade generative AI applications on AWS. It covers foundation model integration, RAG architectures, prompt engineering, agentic AI, vector stores, AI safety and governance, cost optimization, and troubleshooting GenAI solutions using services like Amazon Bedrock, SageMaker, and OpenSearch.
This exam targets developers and ML engineers with 2+ years of production AWS experience and 1+ year building generative AI solutions. You should be comfortable with Amazon Bedrock (Knowledge Bases, Agents, Guardrails, Prompt Management), vector databases, RAG patterns, prompt engineering techniques, and integrating FMs into applications via APIs. It's ideal for AI engineers, full-stack developers working with LLMs, and ML engineers transitioning into generative AI.
Read the full AWS Generative AI Developer Professional certification guide
The AIP-C01 exam has 75 questions (65 scored, 10 unscored) including multiple-choice, multiple-response, ordering, and matching question types. You have 180 minutes and need a scaled score of 750 out of 1000 (approximately 75%) to pass. The exam is available at Pearson VUE and online. The cost is $300 USD.
Recommended experience: 2+ years building production applications on AWS with 1+ year implementing generative AI solutions. Hands-on experience with Amazon Bedrock, RAG patterns, and prompt engineering.
Written in the style of the real exam. Answer each one, then open the answer to check.
Question 1 · Foundation Model Integration, Data Management, and Compliance
A financial services company in Germany runs a generative AI assistant that calls a foundation model in Amazon Bedrock with on-demand throughput in the eu-central-1 Region. During month-end peaks, a growing share of requests fail with throttling errors. Company policy requires that prompts and responses are processed only in AWS Regions in Europe. The company does not want to commit to a fixed amount of model capacity. Which solution will meet these requirements?
Answer: D
Option D is correct: a geographic cross-Region inference profile lets Amazon Bedrock route on-demand requests to other Regions in the same geography, which adds throughput during peaks while keeping processing within that geography and needs no capacity commitment. Option A meets the throughput need but can process requests outside Europe. Option B adds capacity but requires the commitment the company wants to avoid. Option C sends prompts and responses to a US Region, which breaks the residency policy.
Sources: AWS documentation 1, AWS documentation 2
Question 2 · Implementation and Integration
A retailer is building a customer service agent with Amazon Bedrock Agents. The agent must look up order status and issue refunds by calling the company's internal order management REST API, which can be reached only from private subnets in a VPC. Because refunds move money, the agent must ask the customer to explicitly confirm each refund before the refund API is called. The solution must require as little custom orchestration code as possible. Which combination of steps will meet these requirements? (Select TWO.)
Answers: B, C
Options B and C are correct: an action group describes the API operations to the agent through an OpenAPI schema, and a Lambda function attached to the VPC can reach the private API. Turning on user confirmation for the refund operation makes the agent return a confirmation request to the application and invoke the operation only if the customer confirms. Option A gives the agent text about the API but no way to call it. Option D relies on the model following an instruction, which is not a guaranteed control. Option E rebuilds orchestration that Bedrock Agents already provides.
Sources: AWS documentation 1, AWS documentation 2
Question 3 · AI Safety, Security, and Governance
A company has 40 AWS accounts in an organization in AWS Organizations. Teams in many accounts build applications that call foundation models in Amazon Bedrock. The governance team has defined content filters, denied topics, and PII masking that must apply to every model invocation in every account, including accounts that are created later. Application teams must not have to change their code, and the governance team must be able to manage the safeguards centrally. Which solution will meet these requirements with the LEAST operational overhead?
Answer: A
Option A is correct: Amazon Bedrock policies in AWS Organizations enforce a guardrail on all model invocations in the targeted accounts or OUs, including new accounts, without any change to application code; a numeric guardrail version with a resource-based policy lets member accounts apply it while it stays managed centrally. Option B enforces the guardrail, but every application must be changed to pass the guardrail in its requests or its calls fail. Option C depends on every team changing and maintaining its code. Option D only detects missing guardrails after the responses have been returned.
Sources: AWS documentation 1, AWS documentation 2, AWS documentation 3
These sample questions are not part of the AIP-C01 practice exam, so you won't meet them again when you take it.
This practice exam has 75 questions, and exam mode gives you 180 minutes to answer them. The real AIP-C01 exam has 75 questions (65 scored and 10 unscored) in 180 minutes, so a full attempt here is the same length. Practice mode has no timer.
In format, yes: questions are multiple choice (one correct answer) or multiple response (two or more), spread across the 5 exam domains in close to the weights the official exam guide gives them. The real exam can also include ordering and matching questions, which this practice exam does not have. Scoring differs: AWS reports a scaled score from 100 to 1,000 and you pass at 750, while this practice exam shows the raw percentage you got right. For how hard the exam is, whether it is worth it and how long to study, read our AWS Generative AI Developer Professional certification guide.
Aim for 80% or more, consistently, on full-length attempts in exam mode, not just once. That leaves a margin over the 750 out of 1,000 you need on the real AIP-C01 exam, whose questions will all be new to you. Repeat attempts here show the same questions, so judge yourself by first attempts and by whether you can explain each answer. No practice score guarantees a pass.
No. The AIP-C01 practice exam is free and needs no account. Your progress and score history are saved in this browser, so they stay on this device and are lost if you clear your browser data.
Start in practice mode, which tells you after each question whether you were right and explains the answer. When you score well there, switch to exam mode: 180 minutes and no feedback until the end, like test day. Your results break the score down by domain, so you can see which domains to study before your next attempt.
Affiliate links — if you enrol through them, CloudNinja may earn a commission at no extra cost to you.
Studying for more than one certification? All our practice tests are free.