Reserved Instance & Savings Plan Break-Even Calculator

Compare on-demand, reserved-instance, and savings-plan economics with utilization, upfront cost, payback, term savings, and NPV.

A cloud reserved instance or committed-use discount only pays off if you actually use enough of what you committed to — pay a lower committed rate for capacity you don't use, and you can end up worse off than staying on-demand. This calculator compares what you'd actually spend on-demand at your real utilization against the fixed committed cost you're on the hook for regardless of use, then finds the month an upfront commitment fee actually breaks even.

How to use the reserved instance calculator

Enter the on-demand hourly price and the committed hourly price quoted for the same instance or usage, both in dollars.
Add any upfront payment, the hours per month the commitment covers (730 for an always-on month), and your expected utilization as a percentage.
Set the term in whole months — 12 for one year, 36 for three — and a monthly discount rate for the NPV, such as 0.5%.
Read the break-even month, monthly savings, term savings and NPV; a break-even of 'Never' means the commitment loses money at that utilization.

On-demand cost reflects only the hours you use; committed cost doesn't

Monthly on-demand cost is your on-demand hourly rate multiplied by the hours you'd realistically use in a month — that's your monthly hours multiplied by your utilization percentage, not the full clock-hours in a month. Monthly committed cost, by contrast, is the committed hourly rate multiplied by the full monthly hours regardless of your utilization, because that's how a reserved commitment actually bills: you're paying for the capacity you committed to, not the capacity you happened to use that month.

That asymmetry is the whole mechanism behind why utilization matters so much to this decision — at 100% utilization a committed rate almost always wins, but at 30% utilization the on-demand comparison is computed against a much smaller real spend, and the committed rate can lose.

The RI break-even formula

A commitment costs the same whether you use it or not, while on-demand bills only the hours you run. If the committed rate is a discount d off on-demand, you break even when utilization equals 1 − d. Spread over a term, the same idea says a commitment pays for itself once you have run (1 − d) × term months: at 29% off a 12-month term, that's 0.71 × 12, or about 8.5 months of use.

In this calculator the threshold appears as the utilization where monthly savings reach zero, which is the committed rate divided by the on-demand rate. With the defaults of $1.20 on-demand and $0.72 committed that is 60%, and at the default 70% the commitment saves $87.60 a month.

The break-even month the tool reports answers a separate question: how long those monthly savings take to repay the upfront payment. It is upfront ÷ monthly savings, so the default $500 upfront comes back in about 5.7 months and the 12-month term nets $551.20, or about $518 as NPV at 0.5% a month. Drop utilization to 50% and the same offer loses $87.60 a month and never breaks even.

Breakeven, term savings, and NPV are three different answers

Breakeven in months is simply the upfront commitment fee divided by the monthly savings the committed rate produces over on-demand — if the committed rate doesn't actually save money at your utilization level, monthly savings is zero or negative and there is no breakeven month to report.

Term savings totals the monthly savings across the whole commitment term and subtracts the upfront fee, giving a single net dollar figure for the full commitment period. Net present value goes a step further and discounts each month's savings by a monthly discount rate you supply, because a dollar saved in month one is worth more than a dollar saved in month twenty-four — NPV is the number that actually accounts for the time value of the savings, not just their sum.

How to enter No, Partial and All Upfront options

No Upfront: enter 0 as the upfront payment and the effective hourly charge you'll be billed as the committed $/hour.
Partial Upfront: enter the lump sum as the upfront payment and the reduced hourly charge that remains as the committed $/hour.
All Upfront: enter the full prepayment as the upfront payment and 0 as the committed $/hour, since nothing is billed monthly; the break-even month then shows how long your avoided on-demand spend takes to match the prepayment.
Term: use 12 months for a one-year commitment and 36 for three years, and compare the two quotes by NPV rather than by hourly price alone.
Covered hours: leave 730 for an always-on instance, and lower it only if the commitment genuinely covers fewer hours each month.

Reserved Instances vs Savings Plans vs On-Demand

On-demand: no commitment and the highest hourly price, suited to workloads you might stop, resize or move before a term would end.
Reserved Instances: a discount in exchange for committing to a specific instance configuration; on AWS, standard RIs are tied to an instance family and region, while convertible RIs can be exchanged for a smaller discount.
Savings Plans: a commitment to spend a set dollar amount per hour instead of reserving an instance; AWS describes Compute Savings Plans as applying across instance families and regions and to Fargate and Lambda.
Maximum discounts vary by term, payment option and instance type, so take them from AWS's current Reserved Instances and Savings Plans pricing pages rather than from a rule of thumb.
Whichever you pick, the math here is identical: enter the effective committed hourly rate and your realistic utilization.

What gets rejected before any of this runs

Any negative cost, rate, or hour figure; a utilization percentage above 100%; a term of less than one month; or a term that isn't a whole number of months are all rejected outright with an explicit error, because a fractional or zero commitment term has no coherent meaning for this kind of upfront-versus-ongoing comparison.

Where the numbers you enter matter more than the math

Utilization percentage is the single input most likely to be wrong in an optimistic direction — estimate it from actual historical usage data, not from planned or hoped-for usage.
The calculator has no idea what discount tier your actual cloud provider offers for your specific commitment length and payment structure; enter the real quoted committed rate, not an assumed one.
It doesn't model a workload's utilization changing over the commitment term, only a flat rate held constant across every month.
A commitment decision involving meaningful budget should also go through whoever owns cloud cost governance at your organization, not rest solely on this calculator's output.

Questions people ask about Reserved Instance Break-even Calculator

What is the difference between a Savings Plan and a Reserved Instance?: A Reserved Instance discounts a specific instance configuration, while a Savings Plan discounts any eligible usage up to a committed dollar amount per hour, which makes it more flexible. Check AWS's Savings Plans documentation for exactly which services and instance families each plan type covers.
Is a 3-year commitment worth the risk?: Only if you are confident the workload will run at a steady level for all 36 months. Enter the three-year quote with a 36-month term and a realistic discount rate, compare its NPV with the one-year quote, and rerun it at lower utilization to see the downside, since the tool holds utilization flat.
Does this work for Azure reservations or GCP committed use discounts?: Yes, because the math doesn't depend on the provider. Enter the pay-as-you-go hourly price as on-demand, the reserved or committed hourly price, and any prepayment. If your on-demand baseline already includes automatic discounts, use that effective rate rather than the list price.
How do I calculate my RI break-even point?: Divide the committed hourly rate by the on-demand rate to get the utilization you need, for example $0.72 ÷ $1.20 = 60%. For the payback period on an upfront fee, divide the fee by your monthly savings, which is the figure this calculator shows as its break-even month.
Why would a committed-use discount ever cost more than staying on-demand?: Because committed cost is billed on the full capacity you signed up for regardless of how much you actually use, while on-demand cost here reflects only the hours your real utilization implies. At low utilization, the committed rate can end up costing more in total than paying on-demand for just the hours you actually needed.
What does it mean if there's no breakeven month shown?: It means monthly savings from the committed rate, at your entered utilization, is zero or negative — the committed option isn't saving you money at all under those assumptions, so there's no month at which an upfront fee could ever pay itself back.
Why does NPV give a different answer than total term savings?: Term savings simply adds up every month's savings and subtracts the upfront fee, treating a dollar saved in month one the same as a dollar saved in month twenty. NPV discounts later months' savings by the monthly rate you provide, reflecting that money saved sooner is worth more — it's a more conservative, time-adjusted figure.
What utilization percentage should I actually use?: Use your real historical utilization for the resource in question, not a planned or optimistic figure — this is the input most likely to be overstated, and overstating it makes a committed-use discount look better than it will actually perform once you're locked into the term.

Related tools on BrakeTools

EC2 vs. Bare Metal ROI Calculator — Compare amortized total cost of ownership between a cloud VM and bare-metal hardware over a chosen horizon.
Bandwidth Egress Calculator — Compare AWS, GCP, Azure, Cloudflare, and Bunny egress pricing by data volume.
Cloud Storage Price Comparison — Compare S3, R2, Wasabi, and Backblaze B2 storage pricing by GB-month and egress.

Content reviewed 2026-09-27.