Unverified50% confidenceFactExact time
OpenAI引入提示缓存断点机制，允许开发者手动指定缓存断点，缓存命中部分计费通常可打折至原价的10%~25%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/13/2026
First Seen
Valid until: 10/11/2026
Sources
GPT-5.6三款新模型发布：Luna、Terra、Sol主打智能体长任务
rss7/9/2026
Related Claims
VerifiedOpenAI Prompt Caching 可将重复前缀的缓存成本降至原价的10%-25%76% similarUnverifiedAnthropic的Claude API和OpenAI的API均已原生支持提示缓存功能，缓存命中时的输入Token价格通常仅为正常价格的10%-25%75% similarUnverifiedAPI缓存命中机制可使已发送过的输入Token在重复调用时享受折扣价，通常为原价的10%-50%72% similarUnverifiedOpenAI支持自动前缀缓存，折扣幅度约为50%72% similarUnverifiedOpenAI和Anthropic均已推出Prompt Cache功能，可复用KV Cache，对调用方体现为token费用折扣（通常为50%–90%）71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/497880API
curl https://kongchang.com/api/v1/knowledge/claims/497880MCP
get_claim(id=497880)