---
author: "Umesh Malik"
canonical: "https://umesh-malik.com/blog/tag/local-llm"
description: "Explore articles tagged with Local LLM by Umesh Malik — AI Engineer, LLM & GenAI Developer. Learn Local LLM best practices, practical tips, and in-depth guides."
title: "Umesh Malik's Blog - Local LLM Articles | Local LLM Tutorials"
tokens: 275
generator: "scripts/generate-page-markdown.mjs"
---

[← Back to Blog](https://umesh-malik.com/blog)

# Local LLM

2 articles

 [![Muse Glimmer 30B memory ladder showing 55GB at full precision shrinking to under 20GB with 4-bit K-Quant compression](https://umesh-malik.com/blog/run-muse-glimmer-30b-locally-cover.png)

AI Coding Agents & DX • Aug 11, 2026

### Run Muse Glimmer 30B locally: 55GB shrinks to under 20GB

How to run Muse Glimmer 30B locally: the K-Quant setup that fits a single 24GB GPU, the drafter model that triples decode speed, and where it breaks.

8 min read

Read more →](https://umesh-malik.com/blog/run-muse-glimmer-30b-locally)

 [![Layer-by-layer LLM inference streaming one transformer layer at a time from disk onto a 4GB GPU](https://umesh-malik.com/blog/run-70b-llm-on-4gb-gpu-airllm-cover.png)

LLM Engineering • Aug 4, 2026

### Run 70B LLM on 4GB GPU: AirLLM's Real Tradeoff

Run 70B LLM on 4GB GPU hardware with AirLLM's layer-by-layer inference. The VRAM math is real — you just pay for it in disk bandwidth. The honest tradeoff.

11 min read

Read more →](https://umesh-malik.com/blog/run-70b-llm-on-4gb-gpu-airllm)