# Running LLMs on Apple Silicon: a practical guide

> How M1/M2/M3/M4 Macs run local AI, what unified memory means, which models fit your Mac, and how to get started with Metal.

*Source: https://velstech.net/apple-silicon-llm-guide · Updated: 2026-08-30 · Category: AI · Tags: Apple Silicon, Metal, Local AI*

*Markdown version of [Running LLMs on Apple Silicon: a practical guide](https://velstech.net/apple-silicon-llm-guide). [Read the full guide with interactive tools](https://velstech.net/apple-silicon-llm-guide).*
*Also as Markdown: [Hindi](https://velstech.net/apple-silicon-llm-guide.hi.md) · [Tamil](https://velstech.net/apple-silicon-llm-guide.ta.md).*

---

Apple Silicon Macs – M1, M2, M3, and M4 – are quietly some of the best local-AI
machines you can buy. Not because the GPU is the fastest, but because of
**unified memory**: the same RAM is shared by CPU and GPU, which is
exactly what running an LLM wants. A Mac with enough RAM can hold models that would
need an expensive discrete GPU on a PC.

Here's what you need to know about running LLMs on a Mac, what will fit your machine,
and how to get started.

## Why Macs are good at local AI

An LLM's speed is mostly limited by how fast you can feed its weights to the compute
units – that's memory bandwidth. Apple Silicon's unified memory means:

- No VRAM wall. On a PC you're capped by GPU VRAM (12–24 GB typically). On a Mac, the entire system RAM is usable for the model. A 64 GB M-series Mac runs models a 24 GB PC GPU can't.

- Efficient Metal backend. llama.cpp and Ollama have excellent Metal support, so models use the GPU automatically.

- Low power. Even a M-series laptop draws little power running inference, so it's a genuinely practical always-on AI machine.

The catch: **RAM is everything,** and it's soldered. Buy the RAM you'll
want in three years, because you can't upgrade it later.

## What fits your Mac

Model size at 4-bit (Q4_K_M) roughly matches RAM needed, plus the system needs a
few GB for itself:

| Mac RAM | Comfortable model (Q4) | Use case |
| --- | --- | --- |
| 8 GB | 3B–4B | Basic chat, light tasks |
| 16 GB | 7B–9B | Solid daily driver |
| 24 GB | 13B–14B | Serious local AI |
| 32 GB | 27B–30B | Large models, good context |
| 64 GB+ | 70B | Frontier-adjacent |

Memory bandwidth also matters for speed – the M-series Pro/Max chips have more
bandwidth than the base M-series, so a "same RAM, bigger chip" Mac runs faster.

## Getting started

The easiest path is Ollama, which detects Metal automatically. Install it, then:

```
brew install ollama
ollama pull llama3.2:3b
ollama run llama3.2:3b
```

Ollama uses your Mac's GPU out of the box. For more control, build llama.cpp with
Metal and use it directly – same concept as the
[llama.cpp guide](https://velstech.net/llama-cpp-guide), just with the Metal backend.

New to local AI entirely? The [getting
started guide](https://velstech.net/how-to-get-started-local-ai) walks through the whole flow.

## Performance expectations

Expect **5–20 tok/s** decode on an M-series Mac with a model that fits
comfortably – usable for chat, not instant. The exact number depends on your chip,
RAM, model size, and quantization. Compare against real results in the
[benchmark database](https://velstech.net/benchmarks/index) (we test Radeon, but the
decode formula – bandwidth ÷ bytes per token – applies the same way).

Use the [LLM VRAM Calculator](https://velstech.net/llm-vram-calculator) to estimate whether
a model fits your Mac's RAM at your chosen context and quantization.

## Practical tips

- Buy RAM over storage. Storage is cheap to add externally; soldered RAM is forever.

- Close other apps when running big models – the model competes with them for the same unified memory.

- Prefer Q4_K_M to fit bigger models; see the quantization deep dive for the quality trade-off.

- Pro/Max chips run faster at the same RAM due to more memory bandwidth.

## Bottom line

Apple Silicon is one of the most practical local-AI platforms. It is quiet and
low-power. Thanks to unified memory, it can run models that would choke a PC GPU
with half the RAM. A 16 GB Mac is a genuine daily-driver local-AI machine. A
32 GB+ one is a serious workstation. The key decision is buying enough RAM upfront.

## FAQ

**Can a Mac run LLMs?**

Yes – and often better than a PC at the same price, because unified memory means the entire system RAM is available for the model. A 16 GB Mac is a solid local-AI daily driver.

**How much RAM do I need for local AI on a Mac?**

Roughly: 16 GB runs 7B–9B models at Q4, 24 GB runs 13B–14B, 32 GB runs 27B–30B, and 64 GB runs 70B. Buy the RAM you'll want in a few years – it's soldered.

---

*VelsTech – technology explained for everyone. Original: https://velstech.net/apple-silicon-llm-guide*