3 cookbooks for the first Valkey-native semantic cache. Standalone, framework-agnostic, with built-in OpenTelemetry and Prometheus metrics at the cache-operation level.
Install, connect to Valkey, write an embed function, and serve semantically similar prompts from cache - no LLM call needed.
Drop-in caching for LangChain ChatModels via BetterDBSemanticCache, and for Vercel AI SDK models via createSemanticCacheMiddleware.
Threshold tuning, uncertain hit strategies, per-category overrides, TTL management, invalidation, and Prometheus metrics dashboards.
Valkey-native semantic cache. Standalone - works with any LLM client. Built-in OTel tracing and Prometheus metrics.