Separate usage-based inference costs from an AI product’s subscription economics.
Use AI API Cost when you know input and output token volumes and provider rates. Use AI App Profit when you have a blended cost per request, paying-user count, subscription price, and other monthly costs. The cost-per-request model does not automatically import the token estimate; enter a consistent blended cost yourself.
Balance subscription income against inference and hosting.
Output: Estimated operating profit · USD / month
Limitations: All requests belong to paying users. Include free-tier usage and additional expenses in other monthly costs. Excludes taxes.
Estimate token costs with your provider’s own rates.
Output: Estimated API cost · USD / month
Limitations: Enter current provider rates. Caching, batch discounts, tool calls, images, audio, and other charges are excluded.
All prefilled rates are editable examples. Replace them with your actual billing agreements and analytics. Keep the units and reporting period consistent when comparing models. Monthly annualization assumes twelve identical months; it is not a forecast.
Last reviewed: .