Unverified50% confidenceFactExact time
In Claude Code, each model has its own prompt cache; switching models mid-session discards the cache, and the first request after switching re-reads the entire conversation with zero cache hits.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/18/2026
First Seen
Valid until: 1/16/2027
Sources
Related Entities
Related Claims
UnverifiedClaude Code's claude -p (i.e., --print) is a non-interactive mode that passes in the prompt, outputs the model's result, and exits directly, making it suitable for scripted invocation.72% similarUnverifiedClaude Code optimizes through prompt caching, storing recent conversations in a high-speed cache to reduce the cost of repeated reads, but this only affects the unit price and does not reduce the total number of tokens that need to be processed70% similarUnverifiedThe prompt caching mechanism of language models caches previously sent messages, eliminating the need to repeatedly reprocess the same content.69% similarUnverifiedRunning all sessions expiring at the same tick as a single batch allows the model weights to be read only once for the entire group rather than once per session.69% similarUnverifiedCache Read refers to the model directly reading intermediate computation results of already-parsed context from cache during multi-turn conversations or repeated calls, without needing to reprocess.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1057490API
curl https://kongchang.com/api/v1/knowledge/claims/1057490MCP
get_claim(id=1057490)