How Claude Code prompt caching can cut repeat token costs

Claude Code can temporarily save earlier parts of a conversation and reuse them when the same session continues. The already cached part is billed at only 10% of the normal cost, while the newly added part is billed normally. This is called .

Each turn checks whether the beginning of the request matches a recent cache, including the model, , , and full . In per-ing setups, Claude Code uses a default of 5 minutes. If more than 5 minutes pass before the next turn, the whole conversation may be billed at full price again.

A cache write happens when Claude Code saves the current conversation state for a set time, such as 5 minutes or 1 hour.

Key points

  • Claude Code can reuse earlier parts of an ongoing conversation through .
  • Cached prompt content is billed at 10% of the normal cost.
  • The default in per-ing setups is 5 minutes.
  • After a pause longer than 5 minutes, the next turn may lose the cache discount.
  • Long need time management as well as token management.
Read original