AWS instance idle costs
Every AWS instance type costs money whether it's serving traffic or sitting idle overnight — cloud billing doesn't know the difference. Below is what each of the 83 AWS instance types across 18 families costs at its cheapest listed region, running 24/7. Pick a type to see the full region breakdown and what a schedule saves.
c5
C5 is AWS's compute-optimized Intel family — a high vCPU-to-memory ratio built for CPU-bound work like batch processing, video encoding, and CI/CD build fleets.
c6i
C6i is the sixth-generation Intel Ice Lake compute-optimized instance, offering a meaningful throughput bump over C5 for the same CPU-bound profile — batch jobs, encoding pipelines, compute-heavy CI runners.
c7i
C7i is the current-generation Intel Sapphire Rapids compute-optimized family, the fastest per-vCPU option in the C-line for CPU-bound batch, encoding, and simulation workloads.
g4dn
G4dn pairs NVIDIA T4 GPUs with Intel Cascade Lake vCPUs, the workhorse choice for ML inference, small-scale training, video transcoding, and cloud gaming or virtual workstations.
g5
G5 steps up to NVIDIA A10G GPUs, aimed at heavier ML inference, mid-size training, and graphics workloads that outgrow the T4.
g6
G6 uses NVIDIA's current-generation L4 GPUs for the best price-performance in the inference and graphics tier, increasingly the default for new ML-serving and rendering environments.
m5
M5 is AWS's long-standing general-purpose family on Intel Xeon (Skylake/Cascade Lake): a balanced ratio of vCPU to memory that suits application servers, mid-size relational databases, and general backend workloads.
m5a
M5a swaps Intel for AMD EPYC at the same general-purpose vCPU/memory ratio, typically at a modest discount over M5 for comparable throughput.
m6i
M6i is the sixth-generation Intel Ice Lake refresh of the general-purpose line, with higher per-core performance and more consistent latency than M5 at a similar price point.
m7i
M7i is the latest-generation Intel Sapphire Rapids general-purpose instance, with DDR5 memory and improved networking over M6i/M5.
p3
P3 carries NVIDIA V100 GPUs built for serious distributed training and HPC — the kind of instance a team rents by the hour for a training run.
p4d
P4d.24xlarge packs eight NVIDIA A100 GPUs for large-scale model training and is one of the most expensive instances you can run on-demand — over $20/hour even in its cheapest region.
r5
R5 is AWS's memory-optimized Intel family, built for in-memory caches, real-time analytics, and memory-hungry databases that need more RAM per vCPU than general-purpose instances offer.
r6i
R6i is the sixth-generation Intel Ice Lake refresh of the memory-optimized line, offering better memory bandwidth and per-core performance than R5 for the same class of workload — in-memory stores, analytics engines, and large-dataset test environments.
r7i
R7i is the newest Intel Sapphire Rapids memory-optimized instance, pairing DDR5 memory with the highest per-vCPU RAM ratio in the current R-line for large in-memory workloads and memory-bound test environments.
t3
T3 is AWS's third-generation burstable, general-purpose family: a baseline vCPU allocation topped up with CPU credits for short spikes.
t3a
T3a is the AMD EPYC counterpart to T3: the same burstable-credit model, priced roughly 10% below the Intel equivalent for near-identical performance.
t4g
T4g runs on AWS Graviton2 (Arm64) and offers the best price-performance in the burstable family for software that supports Arm — Node, Go, most modern Java and Python stacks.
Stop paying for idle AWS servers
Idlefy runs a free audit of your fleet, then flips dev servers to stopped-by-default: lease one when you need it, and
when the lease ends it shuts itself off — stop, not terminate, persistent disks intact. Timers are set by your team, never
guessed, and Idlefy only manages instances tagged idlefy=enabled — on AWS that boundary is enforced by the IAM policy itself.
Prices are on-demand list price for Linux instances at each type's cheapest listed region and exclude storage, network, data transfer, and any negotiated or volume discounts. Actual bills vary by usage and account-level pricing. Prices as of 2026-07-20.