The price of every thousand API calls at your token sizes.
A single call costs a fraction of a cent — meaningless until you multiply by volume. Cost per 1,000 makes model differences legible. Full workload estimate on the AI app cost estimator.
The Cost per 1,000 Requests Calculator turns your per-request token usage and your model's token prices into the dollar cost of every thousand API calls. You enter four values: input tokens per request, output tokens per request, the input price per 1M tokens, and the output price per 1M tokens. The tool multiplies each token count by its matching per-million rate, adds the input and output sides together to get the cost of a single request, then scales that up by 1,000. The result is a single, comparable number — cost per 1,000 requests — that is far easier to reason about than fractions of a cent per call.
This matters because API bills scale with volume, and a per-request cost that looks trivial becomes significant once you multiply it across real traffic. The key trade-off to watch is that output tokens usually cost several times more than input tokens, so a verbose response can dominate the total even when your prompt is short. Before committing to a model, run your realistic token sizes through the calculator and compare a cheaper model against a pricier one at the same volume; trimming output length or capping response tokens often lowers the per-1,000 cost more than switching providers does.
Cost per request = input tokens × input price + output tokens × output price (per million). Multiply by 1,000 for the per-thousand figure — the unit most API budgets are reasoned in.
Per-request costs are tiny fractions of a cent that are hard to compare. Per-1,000 is the scale where differences between models become obvious and budgetable.