Unverified50% confidenceFactExact time
Mainstream large model API providers typically maintain prompt caches in units of 5 minutes to 1 hour, after which they expire.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/7/2026
First Seen
Valid until: 1/5/2027
Sources
Related Entities
Related Claims
UnverifiedCache entries have a validity period of approximately 5 minutes (default) to 1 hour (extendable), after which they are treated as cache misses.77% similarVerifiedAPI calls to large models typically incur latency of hundreds of milliseconds to several seconds, and response times are unstable due to multiple factors such as network conditions and server load.77% similarUnverifiedThe cache TTL for the subscription version can stay warm for one hour, while the API call mode defaults to expiring after five minutes.77% similarUnverified大模型调用是典型的I/O密集型操作，一次大模型API调用的延迟通常在1-30秒之间73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/984911API
curl https://kongchang.com/api/v1/knowledge/claims/984911MCP
get_claim(id=984911)