OCI block volume cost comes from two dials. The first is capacity, billed per gigabyte per month for the size you provision, not the size you use. The second is performance, set as an elastic level in VPUs per GB, which raises both the guaranteed IOPS and throughput and the price. The four standard points are Lower Cost at 0 VPU, Balanced at 10 VPU, Higher Performance at 20 VPU, and Ultra High Performance from 30 VPU upward. The waste is predictable: most volumes are provisioned a tier or two above what the workload needs, and detached or oversized volumes keep billing in the background. Tune the tier to the real need and clean up the strays, and block storage spend falls with no effect on live workloads.
This is how OCI block volume pricing works, what each tier costs and delivers, where the waste hides, and how auto tune helps.
How is OCI block volume priced?
You pay per gigabyte per month for the capacity you allocate, plus a performance charge that scales with the VPU per GB setting on the volume. Both parts are billed on provisioned size, so a 1 terabyte volume that holds 200 gigabytes still bills for 1 terabyte, and a volume left at a high performance tier bills for that performance whether or not the workload uses it. The charge continues while a volume is detached, which is why orphaned volumes are such a common and quiet source of spend. Boot volumes follow the same model, so an estate with hundreds of instances carries hundreds of boot volumes that are easy to forget.
What do the performance tiers mean?
Performance is elastic: you set a VPU per GB level and OCI guarantees IOPS and throughput that scale with both the VPU setting and the volume size. The four standard tiers cover the range from cheap and slow to fast and expensive.
| Tier | VPU per GB | Profile | Typical fit |
|---|---|---|---|
| Lower Cost | 0 | Lowest IOPS and throughput, lowest price | Archive, cold data, detached volumes, light dev |
| Balanced | 10 | General purpose IOPS and throughput | Most boot and application volumes (default) |
| Higher Performance | 20 | Higher IOPS, latency sensitive | Databases and busy application volumes |
| Ultra High Performance | 30 to 120 | Highest IOPS, scales with VPU | High transaction databases and IO bound workloads |
The performance level is online adjustable, so you can move a volume between tiers without detaching it. That means the right tier is a tuning decision you can revisit, not a choice locked in at creation, which is exactly what makes leaving everything at a high tier wasteful.
Where does the waste hide?
Four categories account for most block volume overspend. Volumes provisioned at Higher or Ultra High Performance that serve workloads with modest IO needs pay for guaranteed performance they never consume. Detached and orphaned volumes, left behind when an instance is terminated, keep billing at whatever tier they were last set to. Oversized boot and data volumes carry capacity that will never fill. And stale manual backups accumulate because nothing deletes them on a schedule. None of these affect a live workload, which is why they survive: they are invisible until someone reads the volume inventory against the bill.
List every block volume with its size, performance tier, and attachment state, then sort by monthly cost. Any detached volume, any volume at Ultra High Performance that is not backing a database, and any volume provisioned far above its used capacity is a candidate. The list almost always pays for the hour it takes to produce.
How does auto tune help?
OCI block volumes support auto tune, which adjusts the performance level automatically based on policy. Detached auto tune drops a volume to Lower Cost while it is unattached, so orphaned and idle volumes stop paying for performance the moment nothing is using them, and the tier restores when the volume is attached again. Performance based auto tune can raise and lower the tier with workload demand within a set range. Enabling detached auto tune across the estate is a near free policy change that removes a whole class of idle spend without any per volume effort.
A Fortune 500 enterprise had standardised every OCI block volume at Higher Performance to be safe, carried dozens of detached volumes from retired instances, and ran no backup retention policy. We retiered application and boot volumes to Balanced where IO metrics allowed, kept Higher and Ultra High Performance only for the databases that needed them, enabled detached auto tune so orphaned volumes fell to Lower Cost, and set a backup retention schedule. Block storage spend dropped materially with no measured performance regression on the live workloads. Figures are verified against billing data and anonymized.
Where to go next
Apply the same tiering logic to unstructured data in object storage tiers on OCI, and make sure your resilience design is not silently doubling storage in disaster recovery on OCI without overspend. The full OCI lever set is the OCI cost optimization guide, and the cross cloud storage view sits in the cross cloud cost optimization guide.
Frequently asked questions
How is OCI block storage priced?
What is a VPU in OCI block volume?
How do you reduce OCI block storage cost?
Take the next step on your cloud spend
We are an independent buyer side advisory that cuts public cloud spend across AWS, Azure, GCP, and OCI. We hold over $2.4B in annual cloud spend under management, take zero provider commissions, and deliver a 31 percent median reduction in the first 90 days. Our guarantee is simple: we reduce your cloud spend or we reimburse our service fee. Pricing is either a Fixed Fee scoped up front or Gainshare, a share of verified savings with no retainer and no risk to you.
Put a defensible number on your cloud spend.
No provider in the room, no published price list. Tell us your footprint and we will scope the savings against your billing data — we reduce your cloud spend or we reimburse our service fee.
The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.