On this site's model pages the cached input price appears separately. Anthropic, for example, charges 0.05x to 0.1x the base input price for cache hits depending on the model.
Frequently asked questions
What does Cached input mean?
Input tokens served from the prompt cache, billed at a reduced rate. On this site's model pages the cached input price appears separately. Anthropic, for example, charges 0.05x to 0.1x the base input price for cache hits depending on the model.