---
title: "Discussion on \"How much VRAM do you actually need to run Llama 3 or Gemma locally?\" | Hashnode"
description: "Discussion on \"How much VRAM do you actually need to run Llama 3 or Gemma locally?\". The weights fit. The KV cache is what OOMs you. Here's the math for Llama 3 and Gemma 2, plus a calculator."
url: "https://hashnode.com/posts/how-much-vram-do-you-actually-need-to-run-llama-3-or-gemma-locally/6a324007a4b114a4a78d6fd8"
---

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

Discussion

# [How much VRAM do you actually need to run Llama 3 or Gemma locally?](https://sathvickollu.hashnode.dev/how-much-vram-do-you-actually-need-to-run-llama-3-or-gemma-locally?utm_source=hashnode&utm_medium=feed)

The weights fit. The KV cache is what OOMs you. Here's the math for Llama 3 and Gemma 2, plus a calculator.

Jun 17 · 6 min read · sathvickollu.hashnode.dev  
Read the article

-   [#ai](https://hashnode.com/tag/ai)
-   [#machine-learning](https://hashnode.com/tag/machine-learning)
-   [#llm](https://hashnode.com/tag/llm)
-   [#local-llm](https://hashnode.com/tag/local-llm)
-   [#gpu](https://hashnode.com/tag/gpu)
-   [#codeswap](https://hashnode.com/tag/codeswap)
-   [#llm-tools](https://hashnode.com/tag/llm-tools)

## Written by

[Sathvic Kollu](https://hashnode.com/@sathvickollu)

@sathvickollu · Bengaluru, India

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

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
