Free tool
LLM API cost calculator
Cost per request is input tokens times the input price plus output tokens times the output price, per million tokens. Thinking tokens are billed as output, and input read from a prompt cache is billed at its cheaper rate. Change any number and the results follow.
The prices start at an example of $2 and $10 per million tokens. Copy your model's real prices from its pricing page: Claude, OpenAI, Gemini.
Cost per request, day and month
- Per request
- $0.007
- Per 1,000 requests
- $7.00
- Per day
- $7.00
- Per 30 days
- $210.00
Per request: $0.004 for input and $0.003 for output (43% output).
Why long chats cost more per turn
A chat sends its whole history with every new message. Each bar is the history re-sent on that turn.
Turn 10 re-sends 1,800 tokens of history; the whole conversation re-sends 9,000, about $0.018 at your input price.
Day 2 of the free Week 1 teaches where these numbers come from, and Day 6 logs the real token counts and cost of every call you make.