LLM cost calculator
Prices differ by provider and change without notice, so this tool does not ship a hardcoded table. You supply the per-million prices; it does the arithmetic and the sanity checks.
$0per month
$0per request
$0per 1,000 requests
0tokens per month
How to read the result
Output tokens usually cost three to five times input tokens, which is why chatty answers get expensive faster than long prompts. If your bill surprises you, it is almost always the output length, not the input.
What this does not model
Caching, batch discounts, and per-provider minimums. Those can change the number by more than half in either direction, so treat this as a ceiling for a naive implementation and re-run it with your cached rate once you know it.