How Much Does Cloud Storage Cost

Cloud providers don’t sell just one kind of storage. They sell a spectrum. And where your data sits on that spectrum shapes your per-GB cost more than anything else. Backup, archive, and active application storage often follow different access and retention patterns, which is why storage tier selection should align closely with workload behavior.

Hot storage is designed for data you access constantly. It carries the highest per-GB rate, but there are no retrieval fees, which makes the total cost predictable. 

AWS S3 Standard comes in at $0.023 per GB-month for the first 50 TB. Google Cloud Standard storage in a multi-region runs $0.026 per GB-month. Azure’s Hot tier for North America sits at $0.0458 per GB, according to Microsoft’s Blob Storage pricing page. Hot tiers make sense for active databases, frequently served assets, or anything that needs to be read without delay or penalty.

These tiers are for data you don’t touch very often, like monthly backups, older logs, and infrequently accessed records. The storage rate drops. Azure Cool runs $0.0230/GB for storage, but pulling data back out adds a retrieval charge of $0.01/GB on top. Azure Cold drops further to $0.0081/GB, with retrieval at $0.03/GB. The tradeoff is straightforward: you pay less to store, but more to access.

Archive storage is for data that rarely moves, such as compliance records, legal holds, and disaster recovery copies you hope to never use. AWS Glacier Deep Archive prices out at $0.00099 per GB-month, roughly $1 per TB-month. Azure’s Archive tier runs $0.0030/GB for storage, but retrieval adds $0.022/GB more. That gap between storage cost and retrieval cost is intentional, and it catches teams off guard when they need the data.

The per-GB rate is the starting number, not the final one. Several other line items quietly inflate storage bills, sometimes more than the storage itself.

Retrieval and egress fees are often the biggest surprise. Pulling data from cool or archive tiers triggers per-GB charges every time, and they stack directly on top of whatever the tier already costs. Azure’s cost estimation documentation breaks this out explicitly, making clear that cooler tiers are only cheap if the data genuinely stays put.

API and operation costs compound fast on certain workloads. Google Cloud’s pricing page shows Class A operations ranging from $0.005 per 1,000 on Standard storage up to $0.05 per 1,000 on Archive, a full 10x jump. For workloads issuing high volumes of read or write requests, operation charges can exceed the storage rate for colder tiers.

Data transfer adds yet another layer. AWS charges separately for inter-region movement and accelerated transfers. Their S3 pricing page includes rates such as $0.04 per GB for certain accelerated transfers. Moving data out of a cloud environment, or between regions, costs money every time it happens, regardless of which tier it came from.

This accumulation of hidden costs is why Flexera’s 2026 State of the Cloud Report found that 29% of IaaS and PaaS spend is wasted. A lot of that traces back to orphaned storage, wrong tier placements, and data movement that nobody budgeted for.

Storage pricing has fine print. Three traps in particular catch teams off guard and erase the savings they expected.

Cheaper tiers require minimum storage commitments, and early deletion still triggers charges for the full period. On AWS, S3 Standard-IA has a 30-day minimum, Glacier Flexible Retrieval requires 90 days, and Glacier Deep Archive locks in 180 days. 

Google Cloud applies 30 days for Nearline, 90 days for Coldline, and 365 days for Archive. Azure mirrors this pattern: 30 days for Cool, 90 for Cold, 180 for Archive. Delete data early, and you pay for the unused portion anyway.

This one surprises teams running workloads with lots of small files. AWS documentation notes that objects under 128 KB may still be billed as 128 KB in certain infrequent-access classes. A workload with millions of tiny files can pay significantly more than the headline per-GB rate implies.

Automated tiering sounds smart, and Azure’s lifecycle documentation encourages moving data between tiers automatically over time. However, the economics only hold when retention and access patterns actually match the tier rules. If data gets moved to Archive but still needs occasional reads, retrieval fees hit every time, and repeated tier moves can trigger early deletion penalties on top.

The same terabyte of data can run anywhere from $1 to $26 per month, depending entirely on how it’s stored and used. Here’s how the math breaks down across common use cases.

The cheapest tier often becomes expensive the moment you actually need the data. That’s the part most cost calculators don’t surface upfront.

Controlling storage spend doesn’t require switching providers. It mostly requires aligning how data is stored with how it’s used and building visibility into what’s happening at the infrastructure level.

Start by matching the tier to the access frequency. Don’t pay hot-tier rates for data that never gets touched, and don’t bury active data in an archive where retrieval fees will hit hard. Set up lifecycle policies to move data automatically, but test retrieval patterns first before full automation.

Monitor API calls and data transfer regularly. Provider tools like AWS Cost Explorer, Azure Cost Management, and Google Cloud Billing surface per-service spend breakdowns. Most teams that see unexpectedly high bills find the problem in API call volumes or unmonitored data movement, not raw storage capacity.

Delete orphaned and duplicate data. Flexera’s 2026 report traces much of the 29% waste figure back to forgotten snapshots, test buckets, and unclaimed volumes. They accumulate fast in active cloud environments and often persist for months before anyone notices.

Where workloads are predictable, look at committed-use discounts. Google Cloud’s pricing page notes that commitments can deliver savings of up to 57% on qualifying services. The same principle applies across providers when usage is consistent enough to forecast.

Understanding cloud storage pricing is one thing. Building the governance model that keeps it controlled month after month is another problem entirely. OTAVA helps organizations design storage strategies that match actual access patterns, automate lifecycle policies correctly, and eliminate the recurring waste that compounds over time. Our managed cloud solutions include proactive tier monitoring, right-sizing recommendations, and more predictable monthly pricing, so retrieval fees and surprise egress charges don’t show up as budget shocks at the end of the quarter. If your current cloud storage costs are harder to explain than they should be, reach out to our team. We’ll review your storage footprint and help you build a cost-controlled, resilient data platform that holds up.