Llama 2 70B Chat on Apple M4 Max (64 GB)

Can the Apple M4 Max (64 GB) run Llama 2 70B Chat?

61 pts

Yes, it runs well

Runs well. At Q3_K_M it needs about 39.6 GB of 48.0 GB usable, with an estimated speed of 10.8 tok/s.

Best quant: Q3_K_M
Generation speed: 10.8 tok/s
Prompt processing: 165 tok/s
First token (1K prompt): 6.2 s
Largest safe context: 4K

Download Llama 2 70B Chat

Recommended for the Apple M4 Max (64 GB): Q3_K_M. 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 llama2:70b-chat-q3_K_M

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 Q3_K_M file (opens in a new tab)

huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF (opens in a new tab)

ollama run hf.co/TheBloke/Llama-2-70B-Chat-GGUF:Q3_K_M

Original weights · Hugging Face

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

huggingface.co/meta-llama/Llama-2-70b-chat-hf (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 (opens in a new tab)

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

Meta’s original Llama 2 repo (now archived).

Links checked 7 days 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.	13.6 / 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.	7.8 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	15.0 / 15
Total	61.4 / 100	61 ptsGood (B) · 61 of 100 points: Model quality 13.6/45, Memory fit 25.0/25, Speed 7.8/15, Memory use 15.0/15

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

Every quantization on the Apple M4 Max (64 GB)

Quant	Memory needed	Available	Speed	Fit	Grade · points	Download
`Q2_K`	33.4 GB	48.0 GB	13.7 tok/s	Runs well	60 ptsGood (B) · 60 of 100 points: Model quality 11.3/45, Memory fit 25.0/25, Speed 8.8/15, Memory use 15.0/15	GGUF (opens in a new tab)
`Q3_K_M`	39.6 GB	48.0 GB	10.8 tok/s	Runs well	61 ptsGood (B) · 61 of 100 points: Model quality 13.6/45, Memory fit 25.0/25, Speed 7.8/15, Memory use 15.0/15	GGUF (opens in a new tab)
`Q4_K_M`	47.9 GB	48.0 GB	8.4 tok/s	Tight fit	61 ptsGood (B) · 61 of 100 points: Model quality 15.4/45, Memory fit 24.0/25, Speed 6.9/15, Memory use 15.0/15	GGUF (opens in a new tab)
`Q5_K_M`	54.8 GB	48.0 GB	5.9 tok/s	Heavy offload	40 ptsPoor (D) · 40 of 100 points: Model quality 15.7/45, Memory fit 4.0/25, Speed 5.7/15, Memory use 15.0/15	GGUF (opens in a new tab)
`Q6_K`	62.3 GB	48.0 GB	—	Won't run	4 ptsWon't run (F) · 4 of 100 points: Model quality 4.8/45, Memory fit 0.0/25, Speed 0.0/15, Memory use 0.0/15
`Q8_0`	78.9 GB	48.0 GB	—	Won't run	4 ptsWon't run (F) · 4 of 100 points: Model quality 4.9/45, Memory fit 0.0/25, Speed 0.0/15, Memory use 0.0/15
`F16`	144 GB	48.0 GB	—	Won't run	4 ptsWon't run (F) · 4 of 100 points: Model quality 4.9/45, Memory fit 0.0/25, Speed 0.0/15, Memory use 0.0/15	Original weights (opens in a new tab)

Memory figures assume 4K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

65 ptsGood (B) · 65 of 100 points: Model quality 15.4/45, Memory fit 25.0/25, Speed 13.4/15, Memory use 11.7/15 NVIDIA RTX PRO 6000 Blackwell Q4_K_M · 31.9 tok/s
65 ptsGood (B) · 65 of 100 points: Model quality 11.3/45, Memory fit 24.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 5090 Q2_K · 52.4 tok/s
60 ptsGood (B) · 60 of 100 points: Model quality 13.6/45, Memory fit 25.0/25, Speed 6.7/15, Memory use 15.0/15 Apple M1 Max (64 GB) Q3_K_M · 7.9 tok/s
59 ptsGood (B) · 59 of 100 points: Model quality 15.4/45, Memory fit 25.0/25, Speed 6.9/15, Memory use 12.4/15 Apple M4 Max (128 GB) Q4_K_M · 8.4 tok/s
57 ptsUsable (C) · 57 of 100 points: Model quality 15.4/45, Memory fit 25.0/25, Speed 8.3/15, Memory use 8.9/15 Apple M2 Ultra (192 GB) Q4_K_M · 12.3 tok/s

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.