Home / Certifications / AIP-C01
Professional 🆕 New Certification

🧠 AIP-C01 - AWS Certified Generative AI Developer Professional

Integrate foundation models into production applications and business workflows using Amazon Bedrock, agents, RAG, and guardrails - with security and cost efficiency.

180 min | 65 questionsPassing: 750/1000

🆕 Newly launched professional-level AI certification

The AWS Certified Generative AI Developer - Professional (AIP-C01) is a professional-level credential for developers who build production generative AI solutions. The beta period ended March 31, 2026, and the exam is now generally available. It sits at the top of the AWS AI track alongside the foundational AI Practitioner (AIF-C01).

Exam Overview

Exam Details

  • Code: AIP-C01
  • Level: Professional
  • Duration: 180 minutes
  • Questions: 65
  • Passing Score: 750/1000
  • Cost: ~$300 USD
  • Validity: 3 years

🎓 Official Resources

👤 Who It's For

Developers who build production generative AI applications. No formal prerequisites, but AWS recommends 2+ years of AWS experience and 1+ year of hands-on generative AI development. Recommended after gaining GenAI experience and ideally the AI Practitioner and/or ML Engineer Associate credentials.

📊 Exam Domains

Domain (representative)Focus
1. Foundation Model Selection & IntegrationChoosing FMs, invoking via Bedrock, embeddings, multimodal
2. Prompt Engineering & Application DesignAdvanced prompting, RAG, agents, orchestration
3. Customization & OptimizationFine-tuning, distillation, evaluation, cost/latency tuning
4. Responsible AI, Security & OperationsGuardrails, privacy, monitoring, deployment, governance

The exam guide defines the authoritative domains and weights - this is a representative summary. Always confirm against the official AIP-C01 exam guide.

📚 Key Topics

Foundation Models & Bedrock

  • Selecting FMs by capability, cost, latency, context window
  • Bedrock InvokeModel / Converse APIs; streaming
  • Embeddings and vector representations
  • Multimodal inputs (text, image)

RAG & Agents

  • Retrieval-Augmented Generation architectures
  • Bedrock Knowledge Bases; vector stores (OpenSearch, Aurora pgvector)
  • Bedrock Agents & AgentCore for tool use and workflows
  • Chunking, re-ranking, grounding strategies

Customization & Evaluation

  • Fine-tuning vs. RAG vs. prompt engineering trade-offs
  • Model distillation; continued pre-training
  • Evaluation: automated metrics, human eval, LLM-as-judge
  • Latency, throughput, and cost optimization (caching, batching)

Responsible AI & Security

  • Bedrock Guardrails: content filtering, PII, denied topics
  • Prompt injection defense; tool-use authorization
  • Data privacy: VPC endpoints, no-train guarantees
  • Monitoring, logging, human-in-the-loop oversight

☁️ Key AWS Services to Know

CategoryServices
Core GenAIAmazon Bedrock (models, Knowledge Bases, Agents, Guardrails, Evaluations), Bedrock AgentCore
Vector / RetrievalOpenSearch Service, Aurora PostgreSQL (pgvector), Kendra
ML platformAmazon SageMaker AI (for custom models / hosting)
App integrationLambda, API Gateway, Step Functions, EventBridge
Security & OpsIAM, KMS, VPC/PrivateLink, CloudWatch, CloudTrail

📋 Study Checklist

Progress0%
  • Read the official AIP-C01 exam guide
  • Invoke foundation models via Bedrock (InvokeModel / Converse)
  • Build a RAG app with Bedrock Knowledge Bases + a vector store
  • Create a Bedrock Agent / AgentCore workflow with tool use
  • Configure Bedrock Guardrails (content filters, PII, denied topics)
  • Compare fine-tuning vs RAG vs prompt engineering
  • Evaluate model quality (automated + human + LLM-as-judge)
  • Optimize latency and cost (caching, batching, model choice)
  • Secure GenAI apps: prompt injection defense, VPC endpoints, IAM
  • Score 75%+ on practice exams before booking
Browse Study Resources →