OpenAI adds cache diagnostics and manual controls to GPT-6 prompt caching
What happened
GPT-6 的提示词缓存现在默认命中率更高，只要 30 分钟内复用相同的前缀内容就能拿到折扣。新上线的仪表盘可以看缓存的命中率走势，诊断工具能告诉你哪次没命中是因为工具定义变了、具体浪费了多少 token（比如一次 tools_changed 就丢了 5,629 个 token）。开发者现在可以手动设缓存断点、在不破坏缓存的前提下调整推理强度，还能提前...
From AI HOT 精选
Coverage
Follow the reports to see the story from different sides.
- AI HOT (Curated Pool)PickOpenAI ships better prompt caching for GPT-6, plus a dashboard and diagnostics
GPT-6 prompt caching now hits more often by default, with discounts for shared prefixes reused within 30 minutes. A new dashboard tracks hit rates and a diagnostics tool pinpoints misses—e.g., a tools_changed reason costing 5,629 tokens. Developers can set explicit cache breakpoints, adjust reasoning effort without breaking cache, and prewarm context to cut latency. GitHub Copilot reports a >50% drop in tokens needing fresh processing; Manus raised cache hit rates from ~85% to >90% in under a week.