Model Cost

Token calculator

What this prompt costs on one call.

Drop in the text you would send. We count it here, then price one model from each family.

Privacy: your prompt stays in this tab while you move between pages. Refreshing clears it. This page never sends the prompt to a server.

Enter a system message, user turn, or any text you would send. Works on a phone keyboard.

Characters0
Words0

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

How to use the token calculator

  1. Enter the prompt you want to measure

    A system message, user turn, document excerpt, or anything you would send on one call.

  2. Read characters and words

    Those two counts update in this tab as you type. The prompt stays on your device.

  3. Pick one model per family and calculate

    Claude, GPT, Gemini, and Grok each get one model. Calculate to see input cost on the chart. GPT tokens are exact; the others are estimates.

  4. Trim the prompt if the cost is high

    Cut repeated instructions, drop unused context, or switch to a cheaper model in the same family.