Check my GPU

on

Can the Apple M2 Pro (32 GB) run Llama 3.2 3B Instruct?

61 ptsGood (B) · 61 of 100 points: Model quality 15.6/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.3/15

Yes, it runs well

Runs well. At Q6_K it needs about 5.9 GB of 21.4 GB usable, with an estimated speed of 45.7 tok/s.

Best quant
Q6_K
Generation speed
45.7 tok/s
Prompt processing
1,313 tok/s
First token (1K prompt)
0.8 s
Largest safe context
128K

Download Llama 3.2 3B Instruct

Recommended for the Apple M2 Pro (32 GB): Q6_K. Every other quantization has its own download link in the table below.

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

ollama run llama3.2:3b-instruct-q6_K

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

Download the Q6_K file (opens in a new tab)

huggingface.co/unsloth/Llama-3.2-3B-Instruct-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Llama-3.2-3B-Instruct-GGUF:Q6_K

Original weights · Hugging Face

Full-precision files from Meta, for fine-tuning, vLLM, Transformers.

huggingface.co/meta-llama/Llama-3.2-3B-Instruct (opens in a new tab)

Free Hugging Face account needed: request access on the model page, and Meta approves it (usually within minutes to a few hours).

Official repo · GitHub

Reference code, inference scripts and docs from Meta. The weights themselves download from Hugging Face.

github.com/meta-llama/llama-models (opens in a new tab)

git clone https://github.com/meta-llama/llama-models

Meta’s official Llama repo: model cards, prompt formats and download tools.

Links checked 1 week ago against Hugging Face, Ollama and GitHub.

How the B grade is scored

Every result gets up to 100 points from four parts. The points decide the letter grade.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.15.6 / 45
Memory fitFull marks when everything sits in GPU or unified memory; fewer when part spills to system RAM.25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.15.0 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.6.3 / 15
Total61.9 / 10061 ptsGood (B) · 61 of 100 points: Model quality 15.6/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.3/15

Grade scale: S 85+ · A 72+ · B 58+ · C 42+ · D 25+ · F under 25 points.

Every quantization on the Apple M2 Pro (32 GB)

Quant Memory needed Available Speed Fit Grade · points Download
Q2_K 4.5 GB 21.4 GB 108 tok/s Runs well 55 ptsUsable (C) · 55 of 100 points: Model quality 11.0/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 4.9/15 GGUF (opens in a new tab)
Q3_K_M 4.8 GB 21.4 GB 85.1 tok/s Runs well 58 ptsGood (B) · 58 of 100 points: Model quality 13.2/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 5.2/15 GGUF (opens in a new tab)
Q4_K_M 5.2 GB 21.4 GB 65.9 tok/s Runs well 60 ptsGood (B) · 60 of 100 points: Model quality 15.0/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 5.6/15 GGUF (opens in a new tab)
Q5_K_M 5.5 GB 21.4 GB 54.4 tok/s Runs well 61 ptsGood (B) · 61 of 100 points: Model quality 15.3/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.0/15 GGUF (opens in a new tab)
Q6_K 5.9 GB 21.4 GB 45.7 tok/s Runs well 61 ptsGood (B) · 61 of 100 points: Model quality 15.6/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 6.3/15 GGUF (opens in a new tab)
Q8_0 6.7 GB 21.4 GB 33.5 tok/s Runs well 61 ptsGood (B) · 61 of 100 points: Model quality 15.8/45, Memory fit 25.0/25, Speed 13.7/15, Memory use 7.2/15 GGUF (opens in a new tab)
F16 9.7 GB 21.4 GB 15.5 tok/s Runs well 60 ptsGood (B) · 60 of 100 points: Model quality 15.8/45, Memory fit 25.0/25, Speed 9.3/15, Memory use 10.4/15 GGUF (opens in a new tab) Original weights (opens in a new tab)

Memory figures assume 8K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.