AI API calculators

Turn the prices, token counts and window sizes you enter into the answer you actually need: what a call costs, how much text fits, and what is left over.

Calculators in this section

Prices are yours, not ours

No page in this section carries a price list. Providers change rates, add tiers and price cached input separately, and a table copied here would be wrong within weeks while still looking authoritative. You paste the rate from your provider’s own pricing page, and the arithmetic is done on your device.

Tokens are counted, words are estimated

A token count is exact only when the model’s own tokenizer produces it. Everything on this site that converts words to tokens is an estimate with the ratio shown and adjustable, because the true ratio depends on the language, the formatting and the tokenizer. Use it to size a budget, not to predict a bill to the cent.

Which one answers your question

For a bill, use the API cost calculator: it charges input and output at their own rates, because providers price them differently and output is usually the expensive half. It also multiplies one call out to a run, which is where a cost that looked like nothing turns into a number worth knowing.

For a budget, use the tokens and words converter, and for a long prompt use the context window calculator, which reserves room for the answer before it tells you the prompt fits. A window that holds your prompt but not the reply is a window that runs out mid-sentence.

Frequently asked questions