MorrowCache: local proxy dedupes paraphrased LLM queries, ~6.7x faster hits

TurnoverSea9119 · reddit · 2026-09-26

A dev open-sourced MorrowCache, a local OpenAI-compatible proxy that cuts duplicate LLM spend. An adjudicator model checks whether an incoming prompt matches a cached answer's intent: HIT reuses the reply and skips the upstream call, MISS calls the model once and stores it, and a flaky judge fails open so requests never get stuck. Judges: cloud Jev (default) or local Kev/Laya/System One — real intent matching, not cosine embedding vibes. Recorded run: a paraphrase went from a 2817ms miss to a 423ms hit, 6.7x faster and mostly free.

Related event: MorrowCache: Open-Source Semantic Cache Proxy Prevents Duplicate LLM Billing(2 posts)→

Original post →

More from coding & agent

coding & agent channel →