AWS pricing
Amazon Web Services runs the largest public cloud, with the deepest catalogue of instance families (T, M, C, R, I, G, X, P, Inf, Trn) and over 30 regions. EC2 list prices are typically the highest of the major providers, but Savings Plans and Spot pricing can drop effective cost by 30 to 90 percent.
- ▸Widest service ecosystem (300+ services beyond compute)
- ▸Best-in-class compliance and enterprise contracts
- ▸Graviton ARM instances offer 20-40% price advantage
- ▸Highest on-demand list prices of the seven
- ▸Egress fees ($0.09/GB after first 100 GB) hurt at scale
- ▸Pricing complexity (region, generation, tenancy multipliers)
Amazon Web Services pricing table
| Provider | Family / instance | vCPU | CPU type | RAM GB | Storage | Net Gbps | Best for | Hourly $ | Monthly $ | Annual $ | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
|
$ |
vCPU
RAM
$
|
$ | Buy | |||||||
| No instances match the current filters. Try widening the vCPU or RAM range. | ||||||||||||
AWS pricing — frequently asked
What is the cheapest AWS server?
The cheapest AWS instance we track is t4g.nano at about $0.0042/hour, and the lowest monthly list price is roughly $3.07/month. Spot/preemptible and committed-use pricing can lower this further.
How is AWS pricing calculated?
Prices shown are USD on-demand list prices for Linux instances, normalised to hourly, monthly (730-hour), and 1-year reserved equivalents so you can compare AWS against other providers at matching vCPU and RAM. We track 45 AWS instances across 9 families.
Is AWS cheaper than AWS?
It depends on the workload. Compare AWS head-to-head against AWS, GCP, Azure, Hetzner, DigitalOcean, Vultr, and Linode at the same vCPU/RAM tier using the comparison links on this page. Lean providers usually win on flat price/performance; hyperscalers win on regions, managed services, and compliance.