Model Cost

API cost calculator

What a month of calls costs.

Enter tokens per request, how many requests run each day, and how many users send them. We scale that to a monthly bill for one model in each family.

Privacy: you type counts, not the prompt. This page never sends the numbers to a server.

From the token calculator, or your own count.

Expected reply length. Leave at 0 to bill the prompt only.

Calls one user (or your already-totaled traffic) makes in a day.

Seats that each make those daily requests. Use 1 if requests per day is already a team total.

Default 30. Change it if you want a shorter or longer window.

Rates last verified 30 August 2026. Catalog rates from OpenRouter (public model list). Not a notarized vendor price sheet. Month length defaults to 30 days. See catalog rates.

How to use the API cost calculator

  1. Enter tokens per request

    Input and output counts for one call. Count a prompt on the token calculator first if you need those numbers.

  2. Add traffic

    Requests per day, users, and days in the month. Leave users at 1 if requests per day is already a team total.

  3. Pick one model per family

    Same usage is priced for Claude, GPT, Gemini, and Grok so you can compare a monthly bill.

  4. Calculate, then adjust

    If the chart is too high, cut tokens, drop requests, or switch to a cheaper model. The figure is a sketch, not an invoice.

Need a token count first?

The token calculator prices one prompt and shows tokens per family.

Open the token calculator