Asking which cloud is cheapest is the wrong question, because the answer changes with your workload. On list price the four are closer than vendors imply, and the real spread opens up through four levers: commitments, enterprise agreements, license benefits, and egress. OCI tends to lead on data egress and on some flexible compute list rates, GCP applies sustained use discounts automatically and lets you commit on spend or resources, AWS offers the deepest and most flexible commitment instruments through Savings Plans and Reserved Instances, and Azure economics swing hardest on Hybrid Benefit and Dev Test pricing. The only honest comparison models your actual usage on each cloud with the discount you would really use, adds your real data transfer, and compares effective annual cost. Sticker rates mislead by ignoring all four levers.
Here is how each cloud prices, where each wins, and how to run a comparison that survives contact with a real bill.
How does each cloud actually price compute?
All four bill compute per second or per hour, but the discount mechanics differ enough to change the answer. AWS discounts steady compute through Savings Plans, which trade a dollar per hour commitment for a lower rate with broad flexibility across instance families, and Reserved Instances, which trade specificity for a slightly deeper discount. Azure offers Reservations and the Azure Savings Plan, and crucially Hybrid Benefit, which lets you bring existing licenses and changes the effective math substantially for Windows and SQL workloads. GCP applies sustained use discounts automatically as a workload runs through the month, and layers Committed Use Discounts that can be spend based or resource based. OCI prices flexible compute shapes that let you dial cores and memory precisely, and its Universal Credits model offers annual flex or pay as you go. Commitment discounts across the four run roughly 20 to 72 percent against on demand in exchange for utilization risk the buyer carries.
Where does each cloud win on price?
No cloud wins everywhere. The table summarises where each tends to have the structural advantage, holding workload constant.
| Cloud | Structural price advantage | Key discount instrument |
|---|---|---|
| AWS | Deepest, most flexible commitment options and the broadest service catalogue to consolidate spend under one agreement | Savings Plans, Reserved Instances, Enterprise Discount Program |
| Azure | Strong economics for Microsoft estates through license reuse and Dev Test pricing for non production | Reservations, Azure Savings Plan, Hybrid Benefit, MACC |
| GCP | Automatic sustained use discounts and flexible spend based commitments that survive instance changes | Committed Use Discounts, sustained use discounts |
| OCI | Materially cheaper egress, precise flexible shapes, and support cost offsets through Support Rewards | Universal Credits, Support Rewards, BYOL database economics |
The practical implication is that a Microsoft heavy estate with idle licenses may find Azure cheapest, an egress heavy data platform may find OCI cheapest, a workload that values automatic discounting without management may favour GCP, and a large diverse estate may favour AWS for the strength of a single consolidated commitment. None of that is visible on a pricing page.
Why does egress change the ranking?
Data transfer is where the four diverge most and where buyers most often get surprised. Moving data out to the internet, and in some cases between regions, is priced per gigabyte, and on the three hyperscalers it is materially more expensive than on OCI, which has positioned cheaper egress as a deliberate differentiator. For a workload that serves large volumes of data to users or replicates across regions, egress can swing the effective comparison by a wide margin even when compute rates look similar. Any price comparison that omits your real traffic is incomplete, and traffic is precisely the line that is hardest to estimate from a pricing page.
Take one real workload, model it on each cloud with the commitment you would actually buy, add your true monthly egress, and include any license you could reuse. If your comparison spreadsheet only has on demand compute rates, it is not a price comparison, it is a sticker comparison, and it will pick the wrong cloud.
How do enterprise agreements move the real price?
On top of per workload discounts sit enterprise agreements that trade a multi year spend commitment for a tier of discount, and they carry risk. The AWS Enterprise Discount Program lowers rates against a committed spend. The Azure MACC carries a shortfall clause, so unspent commitment is still owed. GCP enterprise agreements and Oracle Universal Credits carry the same use it or lose it structure. The discount looks attractive, but the commitment is a liability if your forecast is wrong, so leverage comes from a credible forecast, benchmark data, timing the renewal, and the real option of placing workloads elsewhere. Effective price is set as much in that negotiation as on the pricing page.
A European SaaS company assumed it should consolidate on the cloud with the lowest advertised compute rate. We modelled its real workloads on all four, including its heavy egress and its existing Microsoft licenses. Once Hybrid Benefit, automatic and committed discounts, and true data transfer were applied, the apparent cheapest option fell to third on effective annual cost, and the decision changed. Selecting on modelled cost rather than sticker rate, and negotiating the commitment from a credible forecast, was part of the work that left the estate materially lighter. Figures are verified against billing data and anonymized.
Where this fits in your multicloud strategy
Price comparison is one input to placement, not the whole decision. To see the cross cloud framework, read the cloud cost optimization guide. For the transfer mechanics that swing these comparisons, read multicloud networking costs explained, and for the strategic question of whether to run more than one cloud at all, read consolidate or diversify, the spend question.
Frequently asked questions
Which cloud is cheapest, AWS, Azure, GCP, or OCI?
Why do cloud list prices mislead buyers?
How should you actually compare cloud prices?
Compare on your real numbers, not the sticker
We model your workloads across AWS, Azure, GCP, and OCI on effective cost, independent of any provider and taking zero provider commissions, so placement and commitment decisions rest on evidence. Our guarantee: 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. Start with our playbook.
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.