AWS Cloud Practitioner GuideAWS Cloud Practitioner

AWS Pricing Models Explained: On-Demand, Reserved, Spot, and Savings Plans (CLF-C02)

The AWS Certified Cloud Practitioner (CLF-C02) exam dedicates 12% of its questions to Billing & Pricing. This isn’t just about memorizing discounts—it’s about understanding when to recommend each mode

7 min read
9 sections
Courseiva Study Hub
JA

Reviewed by Johnson Ajibi, MSc IT Security

12+ years in network and security engineering · Founder, JTNetSolutions Limited & Courseiva

Quick answer

The AWS Certified Cloud Practitioner (CLF-C02) exam dedicates 12% of its questions to Billing & Pricing. This isn’t just about memorizing discounts—it’s about understanding when to recommend each mode

Quick answer: On-Demand offers full flexibility with no commitment, Reserved Instances (1-year or 3-year) provide up to 72% savings for steady workloads, Spot Instances slash costs up to 90% for fault-tolerant or batch jobs, and Savings Plans give similar discounts with more flexibility. For the CLF-C02 exam, know the trade-offs: Reserved Instances lock you into specific attributes, while Savings Plans cover instance family usage across regions. Master the AWS Pricing Calculator, Cost Explorer, and Budgets to ace the 12% Billing & Pricing domain.

Why AWS Pricing Models Matter for the CLF-C02 Exam

The AWS Certified Cloud Practitioner (CLF-C02) exam dedicates 12% of its questions to Billing & Pricing. This isn’t just about memorizing discounts—it’s about understanding when to recommend each model. Real-world architects save 30-70% by choosing the right mix, and the exam tests this judgment. You’ll face scenario-based questions where one wrong choice costs a company thousands.

The four core models—On-Demand, Reserved Instances (RI), Spot Instances, and Savings Plans—each serve distinct use cases. Let’s break them down with concrete numbers and exam-ready logic.

On-Demand Instances: Pay-as-You-Go Flexibility

On-Demand is the default: you pay per hour (or per second for some instances) with no upfront commitment. Use it for:

  • Short-term, spiky workloads like development/testing
  • Unpredictable traffic (e.g., a new app with unknown demand)
  • First-time AWS users learning the platform

Cost: No discount. You pay the full list price. For example, a t3.medium Linux instance in US-East-1 costs ~$0.0416/hour ($30.33/month).

Exam tip: On-Demand is the most expensive option but the safest for variable workloads. Never recommend it for steady-state production servers.

Reserved Instances: Commit and Save Big

Reserved Instances require a 1-year or 3-year commitment to an instance family (e.g., m5.large) in a specific region. You save 40-72% vs On-Demand, depending on payment option (All Upfront > Partial Upfront > No Upfront).

Standard vs Convertible RIs

Feature Standard RI Convertible RI
Instance family change No Yes (e.g., m5 to c5)
Scope Region or AZ Region or AZ
Discount Up to 72% Up to 54%
Best for Stable, predictable workloads Workloads with expected changes

When to use each:

  • Standard RI: A database server running 24/7 that won’t change for 3 years. The highest savings.
  • Convertible RI: A growing app that might need to switch from compute-optimized to memory-optimized instances. Lower savings, but flexibility.

Exam question: A company runs a production web server on a t3.medium instance 24/7. They expect consistent traffic for 2 years. Which model minimizes costs?
Answer: Standard Reserved Instance (1-year or 3-year). Avoid On-Demand; Spot is unreliable for production; Convertible RI gives lower savings than needed.

Spot Instances: Max Savings for Transient Workloads

Spot Instances let you bid on unused EC2 capacity—saving up to 90% vs On-Demand. However, AWS can reclaim them with a 2-minute warning when capacity is needed elsewhere.

Ideal use cases:

  • Batch processing (e.g., video encoding, data analysis)
  • Stateless web servers (behind a load balancer)
  • CI/CD pipelines and testing
  • High-performance computing (HPC)

Critical exam point: Spot Instances are not for stateful databases, production workloads, or anything that can’t tolerate interruption. Use them with Spot Fleet or EC2 Auto Scaling to maintain capacity.

Scenario: A data science team runs 1000 short-lived analysis jobs nightly. They can tolerate failures. What model?
Answer: Spot Instances. If a job is interrupted, it restarts on another Spot instance at minimal cost.

Savings Plans: Flexible Commitment Across Services

Savings Plans (Compute or EC2 Instance) offer the same discounts as RIs but with more flexibility. You commit to a spend amount (e.g., $100/hour) for 1 or 3 years, and AWS applies the discount to any eligible usage.

Two types:

  1. Compute Savings Plans – Apply to any EC2 instance, Lambda, or Fargate usage across regions. Up to 66% savings.
  2. EC2 Instance Savings Plans – Apply to a specific instance family in a region (like Standard RIs). Up to 72% savings.

When to choose Savings Plans over RIs:

  • You run multiple instance types (e.g., m5, c5, r5) in the same account
  • You use Lambda or Fargate heavily
  • You want regional flexibility without re-committing

Exam question: A company uses m5.large and c5.xlarge instances across US-East-1 and US-West-2. They want maximum savings with flexibility to change instance families.
Answer: Compute Savings Plans. Covers both families and regions. EC2 Instance Savings Plans would lock them to one family.

Free Tier Mechanics: What You Actually Get

The AWS Free Tier is critical for CLF-C02. It includes:

  • 12 months free (from sign-up): 750 hours/month of t2.micro or t3.micro EC2, 5 GB S3 standard storage, 25 GB DynamoDB, and more.
  • Always free: 1 million Lambda requests/month, 10 GB CloudWatch logs, 25 GB of Amazon S3 storage (for 12 months after first 12).
  • Trial offers: Short-term trials (e.g., 30 days) for services like Amazon Inspector.

Exam trap: The Free Tier does not cover all services. For example, NAT gateways, VPN connections, and dedicated instances are always paid. Also, data transfer out to the internet (over 100 GB/month) incurs charges even in Free Tier.

Scenario: A student launches a t3.small instance to learn AWS. Will it be free?
Answer: No. The Free Tier covers t2.micro and t3.micro only. A t3.small incurs On-Demand charges.

TCO Calculator vs Pricing Calculator vs Cost Explorer

These three tools are distinct, and the exam tests your ability to choose the right one.

Tool Purpose When to Use
AWS Pricing Calculator Estimate costs before building Planning a new architecture
AWS TCO Calculator Compare on-premises vs AWS costs Migration business case
AWS Cost Explorer Analyze actual past spending Budget tracking, anomaly detection
AWS Budgets Set spending thresholds and alerts Real-time cost control

Exam scenario: A CFO wants to see how much the company spent on EC2 last quarter.
Answer: AWS Cost Explorer. It visualizes historical usage and costs by service, region, or tag.

Another scenario: A startup wants to estimate monthly costs for a new serverless app.
Answer: AWS Pricing Calculator. It provides upfront estimates before deployment.

Scenario Questions with Answers

Q1: A batch processing job runs for 4 hours every night. The job can be restarted if interrupted. What’s the cheapest compute option?
A: Spot Instances. They’re fault-tolerant and cost up to 90% less than On-Demand.

Q2: A production database runs 24/7 for 3 years with no expected changes. Which model maximizes savings?
A: Standard Reserved Instance with 3-year, All Upfront payment. Highest discount (~72%).

Q3: A company uses EC2, Lambda, and Fargate across multiple regions. They want a single commitment that covers all compute.
A: Compute Savings Plans. Covers all three services across regions with ~66% savings.

Q4: A developer launches a t3.medium instance for a new project. They expect usage to double in 6 months, but unsure of instance type.
A: On-Demand first. Once usage stabilizes, switch to Convertible RI or Compute Savings Plans.

Q5: An auditor needs to see if any team exceeded their $500/month budget for RDS.
A: AWS Budgets with a cost budget for RDS. Set an alert at 80% threshold.

Key Takeaway for the CLF-C02 Exam

Mastering AWS pricing models isn’t just about memorizing percentages—it’s about matching the model to the workload’s predictability, duration, and fault tolerance. On the exam, you’ll see scenarios where one wrong answer costs a company thousands. Remember:

  • On-Demand = flexibility, highest cost
  • Reserved Instances = steady workloads, max savings with commitment
  • Spot = transient, fault-tolerant jobs, highest risk
  • Savings Plans = flexible commitment across compute services

Use the Pricing Calculator for estimates, Cost Explorer for actuals, and Budgets for alerts. The TCO Calculator is your migration story.

Ready to test your knowledge? Courseiva offers free AWS Cloud Practitioner practice questions that mirror the exam’s Billing & Pricing domain. Start your free practice set now and lock in your CLF-C02 pass.

Practise CLF-C02 questions

Original exam-style practice questions with detailed, explained answers. Track your weak topics and review missed questions before exam day.

Courseiva provides free IT certification practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics. Explore related practice questions for Cisco, CompTIA, Microsoft Azure, AWS, and other certification exams.