AI Token Cost Calculator
Every hosted language model bills by the token, and it bills input and output at different rates — output tokens are usually several times the price of input, because generating text is the expensive direction. So the cost of one request is the input tokens times the input rate plus the output tokens times the output rate. Enter the current rates from your provider and you get the real per-request and total cost rather than an estimate that silently went stale.
How it works
Frequently asked questions
Why are input and output tokens priced separately?
They cost the provider different amounts of compute. The prompt is processed in a single pass, while generated text is produced one token at a time and each new token has to be attended over again — so providers charge a higher rate for output, often three to five times the input rate.
Where do I find the price per million tokens?
On your provider’s pricing page, where rates are almost always quoted per million tokens. Because the headline numbers are small — a few dollars per million — the cost per request looks trivial until you multiply by volume, which is what the total below is for.