www.awsgood.com
On-Demand pricing is a rate applied to running seconds. RI pricing is a contract applied to a matching footprint, and the match happens hourly regardless of what your Auto Scaling group is doing.
That single sentence explains most of the gotchas. An idle RI is not “saving you nothing” — you already paid. On-Demand capacity that sits unused costs nothing at all.
| On-Demand | Standard RI, 1 yr | Convertible RI, 3 yr | |
|---|---|---|---|
| Commitment | None | 1 year | 3 years |
| Billing granularity | Per second | Hourly match | Hourly match |
| Typical discount vs OD | 0% | ~30–40% | ~50–60% |
| Size flexibility (Linux) | N/A | Yes, same family | Yes, same family |
| Change instance family | N/A | No (exchange only, equal value) | Yes, exchange |
| Resale | N/A | RI Marketplace (US sellers) | No |
| Covers Spot / Fargate | N/A | No | No |
Discount percentages vary by region, OS and instance type — pull FixedPrice from the API instead of trusting any table, including this one.
The two numbers to watch after purchase are coverage (share of instance hours covered by RIs) and utilization (share of RI hours actually consumed). Coverage at 95% with utilization at 60% means you bought for a peak that never came back.
First, find what is actually running and — critically — what is not On-Demand.
aws ec2 describe-instances
--filters "Name=instance-state-name,Values=running"
--query 'Reservations[].Instances[].{Type:InstanceType,AZ:Placement.AvailabilityZone,Platform:PlatformDetails,Lifecycle:InstanceLifecycle}'
--output table
`Lifecycle` comes back empty for On-Demand and `spot` for Spot. RIs are invisible to Spot and Fargate usage, so if a third of your fleet is Spot, cut the number you were about to buy by a third before you do anything else.
Next, spend by type over 90 days. Cost Explorer data lags roughly 24 hours, so do not panic when today's purchase doesn't show up.
```bash
aws ce get-cost-and-usage
--time-period Start=2025-06-01,End=2025-09-01
--granularity MONTHLY
--metrics UnblendedCost UsageQuantity
--filter '{"Dimensions":{"Key":"INSTANCE_TYPE","Values":["m6i.large"]}}'
--output table
Then check existing coverage, because buying on top of an already-covered family is how you end up with a utilization report in the 50s.
```bash
aws ce get-reservation-coverage
--time-period Start=2025-08-01,End=2025-09-01
--granularity MONTHLY
--group-by Type=DIMENSION,Key=INSTANCE_TYPE
Anything already above ~85% coverage doesn't need another RI.
## Buying: the copy-paste sequence
Find the offering ID. Duration is in seconds — `31536000` for one year, `94608000` for three.
```bash
aws ec2 describe-reserved-instances-offerings
--instance-type m6i.large
--instance-tenancy default
--product-description "Linux/UNIX"
--offering-class standard
--offering-type "All Upfront"
--filters Name=scope,Values=Region
--duration 31536000
--query 'ReservedInstancesOfferings[].[ReservedInstancesOfferingId,Scope,AvailabilityZone,FixedPrice,UsagePrice]'
--output table
If you skip the `scope` filter you'll get both zonal and regional rows in the same output. Zonal offerings have an AZ populated and reserve capacity in that one AZ; regional offerings leave `AvailabilityZone` blank and follow instances anywhere in the region. Filter the wrong way and `ReservedInstancesOfferings` comes back as an empty `[]`, which reads like a broken command but is just a too-narrow filter.
Always dry-run first. This error is the success case:
```bash
aws ec2 purchase-reserved-instances-offering
--reserved-instances-offering-id 5ec28df5-EXAMPLE-4c1a-9f0e-EXAMPLE
--instance-count 1
--dry-run
```text
An error occurred (DryRunOperation) when calling the PurchaseReservedInstancesOffering operation:
Request would have succeeded, but DryRun flag is set.
Drop `--dry-run` to buy. The response contains a `ReservationId` — store it somewhere; it is not the same identifier as `ReservedInstancesId` and you'll need both later.
## Every flag that matters, and the value I'd pick
| Flag | What it controls | Recommended |
|---|---|---|
| `--instance-type` | Which footprint gets discounted | The family/size with ≥70% steady-state hours |
| `--offering-class` | `standard` vs `convertible` | `standard` unless migrations are likely |
| `--offering-type` | `No`/`Partial`/`All Upfront` | `All Upfront` if cash flow allows |
| `--duration` | Term in seconds | `31536000` for a first purchase |
| `--instance-count` | How many RIs | Floor of steady state, not peak |
| `--product-description` | OS/platform string | `Linux/UNIX` (VPC-suffixed variants also exist) |
| `--instance-tenancy` | Shared vs dedicated | `default` |
| `--filters Name=scope` | Region vs AZ | `Region` |
`Partial Upfront` costs a bit more over the term but keeps cash free; `No Upfront` costs the most and is mostly useful when the commitment itself is the risk you're hedging.
## Verify after 24 hours, and unwind when the number is wrong
Check the reservation exists and note its end date before you forget.
```bash
aws ec2 describe-reserved-instances
--filters Name=state,Values=active
--query 'ReservedInstances[].[ReservedInstancesId,InstanceType,Scope,AvailabilityZone,InstanceCount,Duration,End]'
--output table
Here's where we got burned. Six zonal `m5.xlarge` RIs were bought in `eu-west-1a`, because that's where the instances happened to be on the day of the purchase. Three weeks later an ASG rebalance moved most of the fleet to 1b and 1c. The bill didn't drop, and the coverage report sat at 41% — the zonal RIs simply had nothing to match.
The verification is one grouped call. `AVAILABILITY_ZONE` in the group-by is what exposed it:
```bash
aws ce get-reservation-coverage
--time-period Start=2025-07-01,End=2025-07-22
--granularity DAILY
--group-by Type=DIMENSION,Key=INSTANCE_TYPE Type=DIMENSION,Key=AVAILABILITY_ZONE
Standard RIs can be modified on the spot — scope, AZ, count, network — with no fee and no new term. We flipped all six from zonal to regional, waited for the modification to report `fulfilled`, then re-ran the coverage query over the following week: 41% → 96%, and the on-demand line in Cost Explorer shrank by roughly the amount we'd forecast. `aws ec2 describe-reserved-instances-modifications` is where you confirm the state; it sits at `fulfilled`, not `completed`.
Where an RI genuinely can't be used, standard RIs can be listed on the RI Marketplace (US-based sellers) or exchanged for equal-or-greater value. Convertible RIs cannot be resold, only exchanged — that's the real cost of the flexibility.
## Cheat sheet
| Situation | Buy |
|---|---|
| Steady 24/7 fleet, no migration planned | Standard RI, 1 yr, regional, All Upfront |
| Fleet moves between AZs | Standard RI, regional scope |
| Sizes change within a family | Standard RI (Linux size flexibility) |
| Migration to a new family within 12 months | Convertible RI, or Compute Savings Plan |
| Bursty, <40% monthly uptime | On-Demand |
| Mix of EC2 + Lambda + Fargate | Compute Savings Plan over RI |
| Need capacity guaranteed in one AZ | Zonal RI |
Before you click purchase: confirm coverage is below 85%, confirm the duration in seconds, dry-run the offering ID, and write down the `End` date in the team calendar. Then check coverage + utilization once a week for a month. If coverage is fine but utilization is below 70%, your next purchase should be smaller — not bigger.
If you'd rather have someone else babysit the coverage reports and the renewal dates, that's the kind of thing awsgood.com handles alongside regular cost reviews.