Estimate your monthly LLM API bill. Pick a model, set how many tokens a request uses and how often your users hit it, and see the number. Rates verified June 14, 2026. It is an estimate from your assumptions; a scan turns it into a measured figure from your code.
Model API only. Real apps also pay for retries, system prompts, tool calls, hosting, vector search, and more — usually adding 30-100%+.
This is an estimate from numbers you typed. PrePrice reads your actual code and gives you the real figure, plus what to charge.
Get your real costMonthly model cost = (input tokens per request × input rate + output tokens per request × output rate) × requests per user per month × active users. Token rates are per million tokens and differ for input and output, with output usually several times more expensive. This calculator applies that formula across every model we track.
Roughly 1 token ≈ 4 characters ≈ 0.75 words in English. A short chat turn might be a few hundred tokens; a request that stuffs in retrieved context, a long system prompt, and tools can run several thousand input tokens before the model writes a word. Output length is set by how much you ask the model to produce.
Because it uses numbers you type. Real apps have variable prompt lengths, retries, caching, multiple models, tool calls, and reasoning tokens you can't see from the outside. PrePrice reads your actual code to replace every estimate with a measured figure.
See every model's rates in the AI Cost Index.
PrePrice reads your project, finds every paid service, and computes your real cost per user and what to charge. Free, 2 to 4 minutes.
Start free