MorrowCache: Open-Source Semantic Cache Proxy Prevents Duplicate LLM Billing

MorrowCache is an open-source local OpenAI-compatible proxy that uses a judge model to semantically deduplicate requests before they reach the model, reusing cached answers when intent matches so rephrased queries no longer incur duplicate billing.

2026-09-26 ~ 2026-09-26 · 2 related posts

• MorrowCache proxy dedupes paraphrased prompts: 2817ms miss to 423ms hit — TurnoverSea9119 · 2026-09-26
• MorrowCache: local proxy dedupes paraphrased LLM queries, ~6.7x faster hits — TurnoverSea9119 · 2026-09-26