Input and output tokens are priced very differently (output is 5× input on most models), and a cached system prompt is billed at a fraction of the input rate. This calculator models all three. Prices are list rates as of September 2026 — edit any cell if yours differ.
| Model | $/M in | $/M out | Cache read | Per call | Per month |
|---|
Sources: Anthropic pricing (official), OpenAI and Google list prices as published on their platform pages, checked 2026-09-14. Cache read rates: Anthropic 10% of input; OpenAI and Google 10–25% depending on model — approximated at the vendor's published rate. Long-context surcharges above 200K tokens are not modelled.
Spending more than you'd like? The Claude Token Economy Playbook covers caching, routing and context hygiene that cut bills 40–70%.
Get the Token Playbook →A free tool by Sigma Foundry · no account needed · privacy