DynamoDB Capacity Calculator

Compare DynamoDB on-demand vs provisioned capacity costs. Find the break-even point and choose the most cost-effective mode for your workload.

KB
GB

Provisioned Mode

Required RCUs
50
Required WCUs
100
Provisioned Monthly
$52.20

On-Demand Mode

On-Demand Read Cost
$32.85
On-Demand Write Cost
$328.50
On-Demand Monthly
$361.35

Comparison

Cheaper Mode
Provisioned
Most frequently occurring value
Monthly Savings
$309.16
85.6% less
Storage Cost
$12.50
50 GB @ $0.25/GB
Planning notes, formulas, and examples

About the DynamoDB Capacity Calculator

Amazon DynamoDB offers two capacity modes: on-demand (pay-per-request) and provisioned (pre-allocated throughput). Choosing the wrong mode can mean paying 5โ€“10x more than necessary. This calculator helps you compare both modes side by side and find the break-even point for your workload.

On-demand mode charges per read and write request with no capacity planning required. It's ideal for unpredictable workloads, new tables, or spiky traffic. Provisioned mode requires you to specify read and write capacity units (RCUs and WCUs) in advance, offering lower per-request costs for steady workloads.

By entering your expected read and write volumes, item sizes, and consistency requirements, this calculator shows you exactly which mode is cheaper and by how much. It also factors in DynamoDB storage costs and optional features like auto-scaling and reserved capacity.

When This Page Helps

DynamoDB capacity mode selection is one of the most impactful cost decisions for NoSQL workloads on AWS. On-demand mode can cost 6.5x more per request than provisioned capacity. This calculator quantifies the difference for your specific usage pattern, helping you avoid significant waste on a service that often runs 24/7.

How to Use the Inputs

  1. Enter the average number of read requests per second your application needs.
  2. Enter the average number of write requests per second.
  3. Set the average item size in KB (affects capacity unit consumption).
  4. Choose between eventually consistent and strongly consistent reads.
  5. Enter storage volume in GB for your table data.
  6. Review the on-demand vs provisioned cost comparison.
  7. Check the break-even analysis to determine which mode to use.
Formula used
On-Demand Read Cost = reads_per_second ร— 3600 ร— 730 ร— $0.00000025 ร— RCU_per_read On-Demand Write Cost = writes_per_second ร— 3600 ร— 730 ร— $0.00000125 ร— WCU_per_write Provisioned Cost = (RCU_needed ร— $0.00013 + WCU_needed ร— $0.00065) ร— 730 Storage Cost = GB ร— $0.25/month

Example Calculation

Result: Provisioned: $71.18/mo vs On-Demand: $230.73/mo

With 100 reads/sec on 2 KB items (eventually consistent, 1 RCU per read) and 50 writes/sec (2 WCUs per write for 2 KB items), provisioned mode needs 100 RCUs and 100 WCUs. Provisioned costs $71.18/month, while on-demand costs $230.73/month for the same throughput. Savings: $159.55/month.

Tips & Best Practices

  • Start with on-demand mode for new tables until traffic patterns stabilize.
  • Switch to provisioned mode once you have 2+ weeks of steady usage data.
  • Enable auto-scaling on provisioned tables to handle traffic spikes without over-provisioning.
  • DynamoDB Reserved Capacity offers further discounts for provisioned mode (1 or 3-year terms).
  • Use DAX (DynamoDB Accelerator) to cache reads and reduce RCU consumption.
  • Strongly consistent reads consume 2x the capacity of eventually consistent reads.
  • Design partition keys for even distribution to avoid hot partitions and throttling.

On-Demand vs Provisioned Capacity Deep Dive

On-demand pricing is $1.25 per million write request units and $0.25 per million read request units. Provisioned pricing is $0.00065 per WCU per hour and $0.00013 per RCU per hour. The math clearly shows provisioned is cheaper per-unit, but the key question is utilization: if you provision for peak but average 20% utilization, on-demand may still win.

Auto-Scaling Best Practices

DynamoDB auto-scaling adjusts provisioned capacity based on utilization targets (typically 70%). It responds within minutes, making it suitable for gradual traffic changes but not sudden spikes. Set minimum capacity high enough to handle baseline traffic without throttling, and use a target utilization of 50โ€“70% for a good balance of cost and headroom.

Reserved Capacity for Maximum Savings

For tables with predictable, steady throughput, DynamoDB Reserved Capacity offers significant discounts. A 1-year commitment provides up to 53% savings, while a 3-year commitment saves up to 76%. Reserved capacity applies to provisioned mode only and can cover base throughput while auto-scaling handles peaks.

Sources & Methodology

Last updated:

Frequently Asked Questions

  • One RCU provides one strongly consistent read per second (or two eventually consistent reads) for items up to 4 KB. Larger items consume additional RCUs proportionally.