AIP-C01 Exam Guide: Domains, Cost, Passing Score and a Study Plan [2026 Guide]

The AIP-C01 exam guide published by AWS covers the AWS Certified Generative AI Developer – Professional exam: 75 questions in 180 minutes, a $300 fee, and a passing score of 750 on a scaled range of 100 to 1,000. This AIP-C01 exam guide explains the five domains and their weights, who the exam is built for, how scoring works, and a six-week study plan mapped to the official blueprint.

AIP-C01 is the Professional step at the top of the AWS AI certification path. It sits above the AWS Certified AI Practitioner and the AWS Certified Machine Learning Engineer – Associate. Together, the three steps replaced the retired Machine Learning – Specialty exam. Where the Associate exam tests ML engineering, AIP-C01 tests the GenAI developer role: integrating foundation models into production applications. If you hold the Associate credential, our MLA-C02 exam guide shows how the two blueprints differ.

AIP-C01 Exam Guide at a Glance

AIP-C01 is a Professional-level exam for developers who ship generative AI on AWS. It has 75 questions (65 scored, 10 unscored), runs 180 minutes, costs $300, and needs 750 to pass. Five domains, led by FM Integration at 31 percent of scored content.

Exam fact	Detail
Exam code	AIP-C01
Certification	AWS Certified Generative AI Developer – Professional
Level	Professional
Questions	75 (65 scored, 10 unscored)
Question types	Multiple choice and multiple response
Duration	180 minutes
Passing score	750 on a scaled score of 100 to 1,000
Exam fee	$300
Delivery	Pearson VUE center or online proctored exam
Languages	English, Japanese, Korean, Simplified Chinese
Validity	3 years; recertify

The scoring model matters for strategy. AWS uses a compensatory model: one overall score decides the result, and no single domain can fail a candidate alone. In addition, the 10 unscored questions are unmarked trial items for future exams.

Who This Exam Is For

AWS aims the exam at developers with 2 or more years of experience building production-grade applications on AWS or with open-source technologies, plus general AI/ML or data engineering experience. It also assumes 1 year of hands-on experience implementing generative AI solutions. The recommended AWS knowledge behind the profile is specific:

• Experience with AWS compute, storage, and networking services
• Understanding of AWS security best practices and identity management
• Experience with AWS deployment and infrastructure as code tools
• Familiarity with AWS monitoring and observability services
• Understanding of AWS cost optimization principles

No certification is required before AIP-C01. AWS names four credentials as helpful preparation: AI Practitioner, Solutions Architect – Associate, Machine Learning Engineer – Associate, and Data Engineer – Associate. None of them is a gate. Instead, production experience is the real gate. The blueprint assumes judgment on vector stores with real data, guardrails on live traffic, and cost calls on running systems.

Wait if your generative AI work so far covers demos and tutorials only. Similarly, data science candidates should read the scope section first: model training is out of scope here.

AIP-C01 Exam Guide: The Five Domains and Their Weights

The five domains are weighted as percentages of scored content, the 65 questions that count. Domains 1 and 2 together carry 57 percent of the score. As a result, FM integration and implementation work decide most results.

Domain	Weight	Scored questions (of 65)	Core topics
1. Foundation Model Integration, Data Management, and Compliance	31%	About 20	FM selection, data pipelines, vector stores, retrieval and RAG, prompt governance
2. Implementation and Integration	26%	About 17	Agentic AI, tool integrations, deployment strategies, enterprise integration, FM APIs
3. AI Safety, Security, and Governance	20%	13	Safety controls, data security and privacy, governance, responsible AI
4. Operational Efficiency and Optimization for GenAI Applications	12%	About 8	Cost optimization, performance tuning, monitoring
5. Testing, Validation, and Troubleshooting	11%	About 7	Evaluation systems, troubleshooting

Domain 1 (31%) is the design domain. Its six tasks run from analyzing requirements and designing GenAI solutions to selecting and configuring FMs. They also cover data pipelines, vector store design, and retrieval for FM augmentation. Prompt engineering strategies and governance close the list. In particular, Amazon Bedrock sits at the center.

Domain 2 (26%) is the build domain. Its five tasks cover agentic AI and tool integrations, model deployment strategies, and enterprise integration architectures. FM API integrations and development tooling complete the set. For example, the guide names Bedrock AgentCore, Strands Agents, and AWS Agent Squad in the agentic skills, and Amazon Q Developer in the tooling skills.

Domains 3 to 5: Safety, Operations and Testing

Domain 3 (20%) is the trust domain. Its four tasks cover safety controls, data security and privacy, governance and compliance, and responsible AI principles. Bedrock Guardrails is the flagship service here. In short, one scored question in five sits in this domain.

Domain 4 (12%) covers operations: cost optimization, application performance, and monitoring systems for GenAI applications. Expect token cost, caching, and model choice trade-offs instead of generic cloud monitoring.

Two tasks make up Domain 5 (11%): evaluation systems for GenAI, and troubleshooting GenAI applications. Evaluation covers automated metrics, LLM-as-a-judge, and human evaluation, applied to failures in retrieval, prompts, latency, and cost.

Exam Format, Cost and Scoring

The exam uses two question types only. Multiple choice has one correct response and three distractors. Multiple response offers five or more options. Those items need every correct response selected for credit; partial selections earn nothing. Unanswered questions count as incorrect; nevertheless, guessing carries no penalty. Therefore, no question should ever be left blank.

The fee is $300, the standard AWS price for a Professional exam. Meanwhile, results arrive within five business days in the AWS Certification Account, not on the day. A failed attempt means a 14-day wait before a retake, and the full fee falls due again. The certification stays valid for three years from the pass date. Holders then recertify by passing the current version.

Where AIP-C01 Sits: The Three-Step AWS AI Path

AWS presents its AI certifications as a three-step path. Step 1 is the AWS Certified AI Practitioner (AIF-C01), a Foundational exam on AI and generative AI concepts. The second step is the AWS Certified Machine Learning Engineer – Associate, on building and operating ML workloads in production. AIP-C01, the Professional exam in this guide, is the third and final step. Our AWS certification changes 2026 roundup tracks version switches across the wider catalog.

The routing rule is simple. App developers who build on foundation models aim at AIP-C01. Data and ML engineers start at the Associate step, and beginners start at AI Practitioner. Skipping a step is allowed, because no step is a formal prerequisite. However, the experience profile still applies.

A Six-Week AIP-C01 Exam Guide Study Plan

This plan assumes about 8 to 10 hours a week for a developer who already matches the target profile. Every week pairs blueprint reading with a build, because production decisions stick better after a system breaks once in your own account.

Week	Blueprint focus	Build and practice
1	Domain 1: requirements analysis, FM selection and configuration	Compare three Bedrock models on one real workload; record quality, latency, and cost
2	Domain 1: data pipelines, vector stores, retrieval	Build a RAG assistant with Bedrock Knowledge Bases over your own documents; tune chunking
3	Domain 2: agentic AI, FM APIs, deployment	Build an agent with two tool integrations on AgentCore and Strands Agents; expose it through an API
4	Domain 3: safety, security, governance	Add Guardrails and privacy controls to the earlier builds; attack your own prompts
5	Domains 4 and 5: cost, monitoring, evaluation	Set token budgets and alarms; run an evaluation set with metrics and an LLM judge
6	Full blueprint review	Take the official practice exam on AWS Skill Builder; re-study weak domains; sit a timed rehearsal

AWS sets out the same shape on the certification page. First, start with exam-style questions. Then take the Official Pretest to find weak areas. Finally, finish with the Official Practice Exam. Above all, train to a pace of 2.4 minutes per question, the average the time limit allows.

Exam Traps That Cost Candidates Marks

1. The one-letter mix-up. AIP-C01 (Professional) and AIF-C01 (AI Practitioner, Foundational) differ by one letter and by almost everything else. Check the code before paying.
2. Studying model training. The guide lists model development and training as out of scope. Weeks spent on training mathematics earn nothing here.
3. Hunting for a domain pass mark. There is none. Only the overall 750 matters under the compensatory model. Balance study time and, above all, never skip Domain 1 at 31 percent.
4. Leaving blanks. Unanswered questions score as incorrect, and guessing carries no penalty. The unscored questions are unmarked, so answer all 75.
5. Half-answering multiple response items. Credit requires every correct option. A careless extra guess can zero the question, so judge each option alone.
6. Prompt engineering tunnel vision. Prompt work is one task (1.6) inside one domain. RAG operations, agents, governance, and evaluation carry most marks.
7. Skipping governance as soft content. Domain 3 is 20 percent of the score, about 13 questions. Safety and responsible AI are a fifth of the exam, not a footnote.

What AIP-C01 Does Not Test

The exam guide lists job tasks the target candidate is not expected to perform:

• Model development and training
• Advanced ML techniques
• Data engineering and feature engineering

Budget zero study hours for those three areas. In short, the exam rewards the developer who integrates, secures, and operates foundation models in production, not the researcher who builds them.

Frequently Asked Questions

How much does the AIP-C01 exam cost?

In fact, the fee is $300, the standard AWS price for a Professional-level exam, before any local taxes at booking.

What is the passing score for AIP-C01?

The minimum passing score is 750 on a scaled range of 100 to 1,000. Scoring is compensatory, so no domain carries a separate pass mark.

How many questions are on the AIP-C01 exam?

There are 75 questions in 180 minutes: 65 scored and 10 unscored trial questions, which AWS does not identify during the test.

Do I need AWS AI Practitioner before AIP-C01?

No. AWS requires no prior certification for AIP-C01. It names AI Practitioner, Solutions Architect – Associate, ML Engineer – Associate, and Data Engineer – Associate as helpful preparation.

How long is the AIP-C01 certification valid?

Three years. Holders recertify by passing the current exam version before expiry.

Where can I take the AIP-C01 exam?

At a Pearson VUE test center or online from home. AWS offers it in English, Japanese, Korean, and Simplified Chinese.

When do results arrive, and what happens if I fail?

Results appear in the AWS Certification Account within five business days. After a fail, a candidate waits 14 calendar days to retake and pays the full fee again.

Sources

• AWS Certified Generative AI Developer – Professional certification page: exam overview, cost, duration, delivery, languages, and target candidate.
• Official Exam Guide (AIP-C01) PDF: domains, weights, task statements, question types, and the 750 pass score.
• AWS Certification FAQs: validity and recertification, results timing, and retake rules.