---
title: "Discussion on \"Qwen3-Coder-Next + MXFP8: The 128GB Local LLM That Runs Predictably\" | Hashnode"
description: "Discussion on \"Qwen3-Coder-Next + MXFP8: The 128GB Local LLM That Runs Predictably\". The MXFP8 MLX quant of Qwen3-Coder-Next (80B total, 3B active, 262K context) on a 128GB Mac: LM Studio 0.4.1 setup for Claude Code."
url: "https://hashnode.com/posts/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably/6aaec3c26eadb167d1790c1e"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Qwen3-Coder-Next + MXFP8: The 128GB Local LLM That Runs Predictably](https://leanzero.hashnode.dev/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably?utm_source=hashnode&utm_medium=feed)

The MXFP8 MLX quant of Qwen3-Coder-Next (80B total, 3B active, 262K context) on a 128GB Mac: LM Studio 0.4.1 setup for Claude Code.

Sep 19 · 34 min read · leanzero.hashnode.dev  
Updated Oct 9  
Read the article

-   [#ai-coding](https://hashnode.com/tag/ai-coding)
-   [#local-ai](https://hashnode.com/tag/local-ai)
-   [#mlx](https://hashnode.com/tag/mlx)
-   [#llm](https://hashnode.com/tag/llm)

## Written by

[Mihai\_LeanZero](https://hashnode.com/@mihaileanzero)

@mihaileanzero · Bucharest, Romania

  
[View profile](https://hashnode.com/@mihaileanzero)

Atlassian consultant and Forge developer at LeanZero. We do Data Center to Cloud migrations and build Forge apps, and I write up what I learn at leanzero.net.

I started LeanZero because I saw an industry full of process-heavy companies charging enterprise prices for work that could be done faster, leaner, more honestly. The Atlassian ecosystem shouldn't need a small army and a six-figure budget to get things done right. I hold Atlassian Certified Administration Expert status: ACP-120, ACP-220, ACP-420 and ACP-520, covering Jira, Confluence, Jira Service Management and cloud org admin. Other certs: PSM I, SSM, AWS Cloud Practitioner. I specialize in Forge app development and run an AI-powered agentic workflow I built myself, so I can build across stacks without a dozen specialists. Cloud migrations are my bread and butter, from small teams to orgs with tens of thousands of users. The MCP Doc Processor came from a simple frustration: nothing on the market fit, so I built it. I'm not selling magic pixie dust. I'll tell you honestly what's possible, what's hard, and what you actually need. If you don't need to pay me, I'll tell you that too.

## Responses 2

Markdown supported · ⌘↵ to post

[Brian · AI News](https://hashnode.com/@brianainews)

[Sep 23](https://hashnode.com/posts/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably/6aaec3c26eadb167d1790c1e/comment/6ab4578fd354fba995be0d5e)

Retrieval quality dies when you ignore query rewriting. I expand vague asks into two or three concrete search strings before I hit the vector store.

[Mihai\_LeanZero](https://hashnode.com/@mihaileanzero)

[Sep 24](https://hashnode.com/posts/qwen3-coder-next-mxfp8-the-128gb-local-llm-that-runs-predictably/6aaec3c26eadb167d1790c1e/comment/6ab4578fd354fba995be0d5e#reply-6ab4ba81b6cb82725e24cb59)

Agreed on query rewriting - a vague prompt going straight into a vector store almost always underperforms once you've expanded it into a couple of concrete phrasings first.