On-demand hourly price across a fleet of EC2 instances.
An instance left running 730 hours costs far more than one scheduled for the ~160 working hours it is actually used. Schedule dev fleets and reserve the baseline. Compare Spot on the spot instance savings calculator.
The EC2 Instance Cost Calculator estimates your monthly Amazon EC2 compute bill by multiplying three inputs: the number of instances you run, the hourly price for each instance, and the running hours per month. Multiplying these together turns an abstract per-hour rate into a concrete monthly figure you can plan against. The main drivers are straightforward β adding instances or leaving them running longer raises the total in direct proportion, and the hourly price varies with instance type, size, and region. It isolates raw compute so you can see how each variable moves the bottom line.
The key trade-off to watch is running hours: an instance billed around the clock costs roughly 730 hours a month, but many workloads only need a fraction of that. Shutting down idle or non-production instances outside business hours, or scaling the instance count to actual demand, often cuts the bill more than chasing a cheaper hourly rate. Remember this figure covers compute alone β storage, data transfer, and other services are billed separately, so treat the result as a baseline rather than your full invoice.
On-demand instances bill per second (minimum one minute) at an hourly rate set by instance type and region. A month of always-on is about 730 hours. Spot instances are far cheaper but can be reclaimed; Savings Plans discount steady usage.
Turn off non-production instances outside working hours, right-size over-provisioned types, use Spot for fault-tolerant work, and commit to a Savings Plan for the always-on baseline. Idle and oversized instances are the usual waste.