Llama 3.3 70B Instruct on Apple M3 Ultra (256 GB)

Can the Apple M3 Ultra (256 GB) run Llama 3.3 70B Instruct?

Yes, it runs well

Runs well. At Q4_K_M it needs about 61.7 GB of 192 GB usable, with an estimated speed of 12.3 tok/s.

Best quant: Q4_K_M
Generation speed: 12.3 tok/s
Prompt processing: 249 tok/s
First token (1K prompt): 4.1 s
Largest safe context: 128K

Best app to run it on the Apple M3 Ultra (256 GB)

On macOS

LM Studio

Desktop app · runs on llama.cpp, MLX · use the MLX build

Runs the MLX build, typically up to 10% faster than GGUF for a model with 70.6B active parameters on Apple silicon.
Point-and-click: search, download and chat in one window.

Open Llama 3.3 70B Instruct in LM Studio

Ollama One command with Ollama’s own build of this model.
llama.cpp Usually the first to support new model architectures.

How we choose: the files published for this model, your operating system and GPU, whether it fits fully in memory, and published speed tests (MLX vs llama.cpp on Apple silicon, vLLM vs llama.cpp vs Ollama on NVIDIA). Details.

Download Llama 3.3 70B Instruct

Recommended for the Apple M3 Ultra (256 GB): Q4_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.

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

ollama run llama3.3:70b-instruct-q4_K_M

Quantized files · GGUF

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

Download the Q4_K_M file (opens in a new tab)

huggingface.co/unsloth/Llama-3.3-70B-Instruct-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Llama-3.3-70B-Instruct-GGUF:Q4_K_M

Original weights · Hugging Face

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

huggingface.co/meta-llama/Llama-3.3-70B-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)

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

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

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q4_K_M, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: S 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs well: How well it runs on the Apple M3 Ultra (256 GB): speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	35.1 / 45
Memory fitFull marks when the model fits in the share of unified memory the GPU can use; fewer when it goes past that limit.	25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	8.3 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	7.4 / 15
Match score	75.8 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the Apple M3 Ultra (256 GB)

Quant	Memory needed	Available	Token generation	Runs	Quality	Download
`Q2_K`	46.9 GB	192 GB	20.1 tok/s	Runs well	57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q3_K_M`	53.2 GB	192 GB	15.9 tok/s	Runs well	68Good quality (B) · 68 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q4_K_M`	61.7 GB	192 GB	12.3 tok/s	Runs well	77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	68.8 GB	192 GB	10.1 tok/s	Runs well	79Great quality (A) · 79 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	76.5 GB	192 GB	8.5 tok/s	Slow	81Great quality (A) · 81 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking.	GGUF · folder (opens in a new tab)
`Q8_0`	93.5 GB	192 GB	6.2 tok/s	Slow	82Great quality (A) · 82 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking.	GGUF · folder (opens in a new tab)
`F16`	160 GB	192 GB	2.9 tok/s	Very slow	82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking.	GGUF · folder (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

77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA RTX PRO 6000 Blackwell Q4_K_M · 31.2 tok/s · Runs great
68Good quality (B) · 68 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 Max (64 GB) Q3_K_M · 10.6 tok/s · Runs well (tight fit)
77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. Apple M2 Ultra (192 GB) Q4_K_M · 12.0 tok/s · Runs well
68Good quality (B) · 68 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 Max (128 GB) Q3_K_M · 10.6 tok/s · Runs well
77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. Apple M3 Ultra (512 GB) Q4_K_M · 12.3 tok/s · Runs well
57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5090 Q2_K · 25.2 tok/s · Runs well (offloads to RAM)

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.