---
title: "Discussion on \"Running  Large MoE LLMs on Modest Hardware with FreeToken\" | Hashnode"
description: "Discussion on \"Running  Large MoE LLMs on Modest Hardware with FreeToken\". A new open-source inference engine designed to run MoE models on hardware that doesn't have enough VRAM to hold them"
url: "https://hashnode.com/posts/running-large-moe-llms-on-modest-hardware-with-freetoken/6aa2a6107f57524dc8baf606"
---

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

Discussion

# [Running Large MoE LLMs on Modest Hardware with FreeToken](https://mamonu.hashnode.dev/running-large-moe-llms-on-modest-hardware-with-freetoken?utm_source=hashnode&utm_medium=feed)

A new open-source inference engine designed to run MoE models on hardware that doesn't have enough VRAM to hold them

Sep 10 · 15 min read · mamonu.hashnode.dev  
Read the article

-   [#local-ai-models](https://hashnode.com/tag/local-ai-models)
-   [#low-end-local-ai](https://hashnode.com/tag/low-end-local-ai)
-   [#gpt-oss-20b](https://hashnode.com/tag/gpt-oss-20b)
-   [#local-llm](https://hashnode.com/tag/local-llm)
-   [#mixture-of-experts](https://hashnode.com/tag/mixture-of-experts)

## Written by

[mamonu](https://hashnode.com/@mamonu)

@mamonu · London, UK

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

quantised braiiins

## Responses

Markdown supported · ⌘↵ to post

No responses yet.