1
PrismLib – semantic LLM cache and cluster mesh that cuts token spend
开源语义缓存神器，91-96%命中率，集群token消耗直降76%，延迟减少98.6%！
Article URL: https://github.com/insightitsGit/prismlib Comments URL: https://news.ycombinator.com/item?id=48693843 Points: 2 # Comments: 0