LMCache
Supercharge Your LLM with the Fastest KV Cache Layer
Overview
LMCache is an open-source KV cache engine that accelerates LLM serving with vLLM and SGLang by storing and reusing KV caches across requests. It delivers higher throughput, lower latency, and drastically reduced costs. LMCache Server introduces a production-grade, distributed KV cache service with a unified interface, prefix caching, semantic chunking, and elastic multi-GPU deployment, making long-context and multi-turn workloads dramatically faster and cheaper.
Requirements
Min vCPU
2
Min RAM
4096 MB
Min Disk
10 GB
Rec vCPU
4
Rec RAM
8192 MB
Rec Disk
20 GB
Recommended VPS
Hostinger · KVM 4
4 vCPU · 16384 MB · 200 GB
Hostinger · KVM 4
4 vCPU · 16384 MB · 200 GB
Hostinger · KVM 8
8 vCPU · 32256 MB · 400 GB
Affiliate disclosure
Docker Compose
# Generated by Run This Ai — docker-compose.yml
services:
lmcache:
image: https://github.com/LMCache/LMCache/blob/dev/docker/Dockerfile
restart: unless-stopped
ports:
- 8080:8080
volumes:
- ./data/lmcache:/data
Guides & articles
How to Deploy LMCache with vLLM: Step-by-Step Tutorial
A hands-on tutorial for installing LMCache, enabling it in vLLM, running the distributed LMCache Server, and verifying time-to-first-token improvements.
Aug 25, 2026
LMCache: The Fastest KV Cache Layer for LLM Serving
LMCache is an open-source KV cache engine that accelerates vLLM and SGLang by reusing KV caches across requests — boosting throughput, cutting latency, and slashing LLM serving costs.
Aug 25, 2026