# LLM for Mac

> Which local model fits which Mac. Specs are cited to Apple, model cards, and vendor docs. Unknown stays unknown.

## Primary

- [LLM Info](https://llmformac.com/llm-info): Machine-readable description of LLM for Mac for AI assistants: what the site is, how to fetch it, and how to cite it.
- [Methodology](https://llmformac.com/methodology): The memory-fit formula used on LLM for Mac: weight bytes, KV cache bytes, and a labeled overhead assumption.
- [About](https://llmformac.com/about): LLM for Mac explains which local model fits which Mac, using Apple spec pages, model cards, and vendor docs.

## Key pages

- [How Much RAM Do You Need to Run a Local LLM on a Mac?](https://llmformac.com/how-much-ram-for-local-llm): Unified memory for a local model is weights plus KV cache plus a labeled overhead, computed from the model card and config.json.
- [Best Local LLM for Your Mac, by RAM Tier](https://llmformac.com/best-local-llm-for-mac): Picks by installed unified memory use the largest cited Qwen2.5 model whose 4-bit total is under that memory. Quality scores are unknown.
- [What Fits My RAM? Local LLM Picker for Mac](https://llmformac.com/what-fits-my-ram): Choose an installed memory size and the table keeps Qwen2.5 rows whose 4-bit or 8-bit total is under that size.
- [MLX vs GGUF on Mac: Which Model Format Should You Use?](https://llmformac.com/mlx-vs-gguf): MLX is Apple's array framework and weight format. GGUF is the llama.cpp file format. Pick the format your app documents.
- [LM Studio vs Ollama on Mac: Which Should You Use?](https://llmformac.com/lm-studio-vs-ollama-mac): LM Studio documents a desktop app with GGUF and MLX on Apple silicon. Ollama documents a macOS app, a CLI, and a local API.
- [Mac Studio vs MacBook Pro vs Mac mini for Local AI](https://llmformac.com/which-mac-for-local-llm): Compare current Mac Studio, MacBook Pro, and Mac mini configs by the unified memory and bandwidth Apple lists.
- [How to Run Llama, Qwen and Mistral Locally on a Mac](https://llmformac.com/guides/run-llm-on-mac): Install Ollama, LM Studio, mlx-lm, or llama.cpp on a Mac using the commands their docs publish, then pick a model that fits.
- [Can a MacBook Air Run a Local LLM?](https://llmformac.com/can-macbook-air-run-llm): A current MacBook Air can run a local model when the model's total is under its 16, 24, or 32 GB of unified memory.

## Guides
Install notes live in this folder.
- [How to Run Llama, Qwen and Mistral Locally on a Mac](https://llmformac.com/guides/run-llm-on-mac): Install Ollama, LM Studio, mlx-lm, or llama.cpp on a Mac using the commands their docs publish, then pick a model that fits.

## About

Owner: LLM for Mac
Standard followed: https://llmbestpractices.com
Updated: 2026-10-12
Sister index: https://llmhardwareguide.com/llms.txt
IndexNow key: https://llmformac.com/910146e3050d9f8058190872277e6d4e.txt