Server Cost Per Request
Calculator

It divides your total server infrastructure cost over a period by the total number of requests handled in that same period, giving you the average cost per request. This is a useful unit-economics metric for understanding how efficiently your infrastructure scales with traffic.

Total spend alone doesn't tell you whether costs are growing faster or slower than usage. Cost per request that trends downward as traffic grows suggests good economies of scale, while a rising cost per request even as traffic grows can signal an efficiency problem worth investigating.

Include all relevant infrastructure costs for the period: compute (servers or cloud instances), storage, bandwidth/data transfer, and any managed services directly tied to serving those requests. Leaving out a major cost category will understate the true cost per request.

Common causes include over-provisioned or idle infrastructure that costs money without serving proportional traffic, expensive database queries or inefficient code paths, or a mismatch between reserved capacity and actual demand. Comparing cost per request across different services or endpoints can help pinpoint where the inefficiency lies.

If you know your cost per request, you can project future infrastructure costs by multiplying it by expected future request volume, and you can check whether your pricing model per user or per transaction comfortably covers the underlying infrastructure cost.