OCI Object Storage offers three tiers: Standard for hot data, Infrequent Access for data read occasionally, and Archive for cold data you almost never touch. Archive holds bytes at roughly a tenth of the Standard rate, on the order of 0.0026 US dollars per GB per month as an indicative list figure, but it adds a retrieval fee per GB restored, a restore wait of about one hour, and a minimum retention of 90 days during which early deletion is still billed. The tier saves money only when the data is truly cold and you can predict that, and OCI’s materially cheaper egress means restoring it costs less to pull out than on the hyperscalers.
Here is how to decide what belongs in archive and what the move actually saves.
What do the three OCI storage tiers cost?
The tiers trade storage rate against access cost and latency. Standard is the most expensive to hold and free and instant to read. Infrequent Access sits in the middle: a lower storage rate with a retrieval fee per GB and instant access. Archive is the cheapest to hold and the most expensive and slowest to read. The whole decision is a bet on access frequency: every read you do not make is saved storage, and every read you do make on archived data carries a fee and a delay.
| Tier | Storage rate | Retrieval | Access latency | Best for |
|---|---|---|---|---|
| Standard | Highest | Free | Instant | Hot, frequently read data |
| Infrequent Access | Lower | Fee per GB | Instant | Read a few times a year |
| Archive | Lowest, near 0.0026 per GB | Fee per GB, plus restore | About one hour | Cold data, rarely read |
When does archiving actually save money?
Archive pays when the storage saving over the data’s lifetime exceeds the expected retrieval cost. Work it as a simple break even: if Archive saves a few tenths of a cent per GB per month against Standard, a terabyte held for a year saves a meaningful sum, but a single full restore of that terabyte plus its egress can claw back months of that saving. So the rule is access frequency. Data you expect to read once a year or less, and can wait an hour for, belongs in archive. Data you might pull on a busy quarter end does not, because the retrieval fees and the restore latency will cost you more than the cheaper shelf.
What traps catch buyers on OCI Archive?
Three. First, the 90 day minimum retention: delete or overwrite archived data before 90 days and you are still billed as if it stayed, so do not archive churny data. Second, the restore wait: archive is not a backup you can read on demand, so anything with a recovery time objective shorter than an hour does not belong there. Third, retrieval at scale: a compliance pull or a legal hold that restores a large slice at once turns a cheap shelf into a real bill, which is exactly where OCI’s lower egress softens the blow relative to the hyperscalers but does not erase it. Model the worst case read, not the happy path.
A European SaaS company kept seven years of raw event logs on Standard object storage for compliance, about 300 TB growing monthly, almost never read. We applied a lifecycle policy moving objects older than 90 days to Archive and kept the most recent quarter on Standard for active investigation. The cold tier cut the storage line on that data by roughly 88 percent at indicative rates. We budgeted an annual retrieval allowance for audits and confirmed restore latency met the compliance recovery window. Figures are verified against billing data and anonymised.
How do you automate tiering safely?
Use Object Storage lifecycle policies to move objects to Archive after an age threshold rather than tiering by hand, and choose the threshold from real access logs, not a guess. Keep a hot window long enough to cover your normal investigation horizon, then archive. Tag data by retention class so the policy is auditable, and budget a deliberate retrieval allowance for the audits and legal holds you can foresee. This sits alongside object storage tiers on OCI and why OCI egress pricing changes the math, inside the OCI cost optimization guide.
Frequently asked questions
Tier your cold data without the retrieval surprise
We model your real access patterns, set lifecycle policies that move only genuinely cold data to Archive, and budget the retrievals you can foresee so the saving is real net of fees. It is the same discipline behind the 31 percent median reduction we see in the first 90 days, with zero provider commissions. Our guarantee: we reduce your cloud spend or we reimburse our service fee.
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.