Azure pricing is a toolkit, not a single rate: pay as you go covers variable demand, Reservations and the Azure savings plan for compute discount a steady base by roughly up to 72 percent for three year terms, spot pricing covers interruptible work at deep discounts, and Azure Hybrid Benefit and Dev Test pricing change the math further where you qualify. The buyer job is to map each workload to the right instrument, covering the durable floor with commitments sized to a defensible forecast and leaving the variable peak on demand, rather than chasing the deepest headline discount. Getting this mix right, with high commitment utilization and the discount programs you are entitled to, is where most of the achievable Azure savings live.
Here is each model, what it is for, and how to combine them without overcommitting.
What are the Azure pricing models, in plain terms?
Pay as you go is the on demand rate: full flexibility, highest unit price, paid for what you use. Reservations commit to a specific resource, a virtual machine size in a region, a database tier, or capacity for services like storage and Cosmos DB, for one or three years, in exchange for a deep discount. The Azure savings plan for compute commits instead to a fixed hourly dollar amount of eligible compute spend for one or three years, applying automatically across regions and sizes for more flexibility at a slightly smaller discount than Reservations. Spot pricing sells spare capacity at steep discounts for workloads that tolerate eviction. On top sit discount programs: Azure Hybrid Benefit, which lets you apply Windows Server and SQL Server licenses you already own, and Dev Test pricing for qualifying non production subscriptions.
Which model fits which workload?
| Model | Indicative discount | Best for |
|---|---|---|
| Pay as you go | None, baseline rate | Variable, short lived, or unpredictable demand |
| Reservations | Roughly up to 72 percent (3 year) | Stable, known resources that will not move |
| Azure savings plan for compute | Slightly less than Reservations | An evolving compute fleet across sizes and regions |
| Spot | Large, capacity dependent | Interruptible batch, testing, stateless work |
| Azure Hybrid Benefit | Further savings on licensing | Workloads where you own eligible licenses |
Discounts are indicative and depend on term, region, and resource; verify against current Azure pricing.
How do you combine them without overcommitting?
Layer the instruments against a demand profile. Find the steady floor that runs reliably, then split it: stable, fixed resources like a production database tier carry Reservations for the deepest discount, while the broader evolving compute base carries the Azure savings plan for flexibility as sizes and regions shift. Variable demand above the floor stays on pay as you go. Interruptible work moves to spot. Across all of it, apply Azure Hybrid Benefit wherever you hold eligible licenses, since it stacks with Reservations and the savings plan. Size every commitment to a forecast you can defend and aim for high utilization, because an unused Azure commitment, like any other, is still owed. Stagger terms so renewals do not all land at once.
A worked example
A European SaaS company ran almost its entire Azure estate on pay as you go, reasoning that flexibility was safer than commitment. Profiling showed a large, stable compute floor that had run flat for over a year, plus production SQL databases on fixed tiers, and a fleet of Windows Server virtual machines whose licenses the company already owned. We put the fixed databases on three year Reservations, covered the evolving compute floor with an Azure savings plan, applied Azure Hybrid Benefit across the eligible Windows estate, and moved batch jobs to spot. The blended effect cut Azure compute and database spend substantially while leaving genuine variability on pay as you go, part of a program that left the company materially lighter on cloud spend. Figures are verified against billing data and anonymised.
Frequently asked questions
What are the main Azure pricing models?
What is the difference between Azure Reservations and the Azure savings plan?
How much can Azure pricing models save?
Map your Azure estate to the right models
We match each Azure workload to the pricing model it justifies and size commitments to a forecast you can defend, as an independent advisory that takes zero provider commissions and answers only to you. Our guarantee: we reduce your cloud spend or we reimburse our service fee, on a Fixed Fee or a no risk Gainshare basis. Download the Azure MACC guide, read the deeper Azure cost optimization guide, and compare instruments in Azure savings plan versus Reservations.
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.