Unclear Which models the shorter 5-minute cache can apply to is not known.
The prompt cache is Anthropic's servers holding on to the earlier part of a conversation for a while, so that it does not have to be processed again on every turn. Claude Code decides how long that cache should last. Subscribers on the allowlist for the longer lifetime used to get 1 hour in every case.
Claude Code now also looks at the model the main conversation is using. If that model is in a particular set and a server-side switch is on, the lifetime is cut to 5 minutes. The switch is off unless it is turned on remotely, and with the switch off the 1-hour lifetime stays as before.
Anthropic can shorten the cache lifetime for particular models without shipping a new Claude Code release. A shorter cache means a long pause between messages is more likely to make Claude reprocess the whole conversation.
tengu_tidy_frost Not enough to sayNothing here resolved what this flag was doing on this version, so nothing here should be read as on or off.
This account: no value returned · anonymous baseline: no value returned · compiled default in v2.1.293: off
These values were read against a different version of Claude Code, so treat them as the nearest reading available instead of one taken on this release.
tengu_prompt_cache_1h_config Not enough to sayNothing here resolved what this flag was doing on this version, so nothing here should be read as on or off.
This account: no value returned · anonymous baseline: no value returned · compiled default in v2.1.293: not a boolean we can read
These values were read against a different version of Claude Code, so treat them as the nearest reading available instead of one taken on this release.
Read once, for one account on one subscription tier, against v2.1.293. It isn't a statement about your account. What a flag value here can and cannot tell you
Which models the shorter 5-minute cache can apply to is not known.
New in this build: tengu_tidy_frost