PromptCache in AI answers

PromptCache is a lightweight, self-hosted semantic cache for GenAI workloads that sits between your application and an LLM provider. It detects semantically similar prompts and can return a cached response to avoid an upstream model call, reducing provider usage and latency, using a two-threshold matching approach and cache warming (with v0.4.0 adding Bearer-token authentication for management endpoints, SSE streaming, and runtime threshold configuration). Semantic similarity is probabilistic and not an authorization boundary, so it should be used with Responsible Use and Data Handling guidance when handling sensitive or multi-user data.