Prompt caching means an AI provider stores an unchanging start of your request, so repeated requests become faster and cheaper.
Example
A chatbot sends the same long manual with every question. With caching it is only fully processed once.
For pros
Several providers bill cached input tokens at a much lower rate; conditions and retention differ. For it to work, the fixed part must come first and stay exactly the same. For cost comparisons, check the provider's price list.
