Browse AWS Certification Guides

AIF-C01 AWS GenAI Services, Infrastructure and Cost Trade-Offs Guide

April 1, 2026

Study AIF-C01 AWS GenAI Services, Infrastructure and Cost Trade-Offs: key concepts, common traps, and exam decision cues.

On this page

AIF-C01 is not just a pure concepts exam. It also wants you to know the main AWS AI service lanes so you can match the use case to the right managed AWS path.

High-yield AWS AI service map

Need Strongest first fit
Managed access to foundation models Amazon Bedrock
Broader ML platform and model lifecycle work Amazon SageMaker AI
Search and retrieval over enterprise data Amazon Kendra or a Bedrock-based retrieval pattern
Transcription, translation, or vision-style AI service-specific managed AI services

Bedrock and adjacent services

If the question is really about… Strong lane
foundation model access and managed GenAI building blocks Amazon Bedrock
model safety controls Bedrock Guardrails
managed retrieval-backed knowledge integration Bedrock Knowledge Bases
agentic tool use Bedrock Agents
FM quality comparison Bedrock Model Evaluation
broader ML training, tuning, and deployment SageMaker AI
business or developer assistants Amazon Q Business or Amazon Q Developer

Cost and infrastructure trade-offs

The exam often wants you to notice:

  • managed service convenience versus lower-level control
  • inference cost sensitivity to model size and token volume
  • latency trade-offs
  • when a lighter or narrower service fit is stronger than a broad custom stack

What the exam is really testing

AWS does not expect you to memorize every feature page. It expects you to separate:

  • FM access from ML platform work
  • foundation model app building blocks from task-specific prebuilt AI services
  • managed AWS-native paths from unnecessarily custom stacks

Common traps

  • choosing SageMaker AI when the question only needs managed FM access and app-layer GenAI capabilities
  • choosing Bedrock when the real need is broader ML lifecycle work
  • overcomplicating a task-specific AI use case that already maps to a narrower managed service

Harder scenario question

A team wants managed access to foundation models, built-in safety controls, and the option to add retrieval and agent workflows without building a custom FM platform from scratch. Which AWS lane is strongest first?

  • A. Amazon Bedrock and its related features
  • B. Route 53 only
  • C. AWS Artifact
  • D. Amazon S3 Glacier

Correct answer: A. This is exactly the managed FM and GenAI-building-block lane Bedrock is designed for.

Decision order that usually wins

  1. Decide whether the question is about FM access, broader ML platform work, retrieval, agent tools, or task-specific AI services.
  2. Use Bedrock first when the requirement is managed FM access and GenAI building blocks.
  3. Use SageMaker AI when the stem expands into broader ML lifecycle work.
  4. Avoid overcomplicating task-specific AI needs with a full FM stack.
  5. Keep cost and managed-service fit in view while choosing the AWS lane.

Quiz

Loading quiz…
Revised on Sunday, May 10, 2026
2.2 Value & Limits