NuExtract 2.0 4B on Apple M2 Max (32 GB)

Can the Apple M2 Max (32 GB) run NuExtract 2.0 4B?

Yes, it runs well

Runs well. At Q5_K_M it needs about 5.1 GB of 21.4 GB usable and makes one decision in about 376 ms.

Best quant: Q5_K_M
Time per decision: 376 ms
Decisions per second: 2.7/s
Input assumed: 256 tokens
Returns: Labeled spans or fields

Best app to run it on the Apple M2 Max (32 GB)

On macOS

LM Studio

Desktop app · runs on llama.cpp, MLX

Point-and-click: search, download and chat in one window.

Open NuExtract 2.0 4B in LM Studio

llama.cpp Usually the first to support new model architectures.
Ollama Runs the Hugging Face GGUF with ollama run hf.co/…

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 NuExtract 2.0 4B

Recommended for the Apple M2 Max (32 GB): Q5_K_M. Every other quantization has its own download link in the table below.

Quantized files · GGUF

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

Download the Q5_K_M file (opens in a new tab)

huggingface.co/numind/NuExtract-2.0-4B-GGUF (opens in a new tab)

ollama run hf.co/numind/NuExtract-2.0-4B-GGUF:Q5_K_M

Original weights · Hugging Face

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

huggingface.co/numind/NuExtract-2.0-4B (opens in a new tab)

No sign-in needed.

Links checked 2 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q5_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: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs well: How well it runs on the Apple M2 Max (32 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.

PrecisionEditorial precision tier from the model’s published benchmarks, scaled by the quantization or precision used.	33.2 / 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
SpeedTime for one decision against a 100 ms target: reading the input plus any tokens the model writes.	7.7 / 15
ClarityHow easy the answer is to use in code: a probability or score earns the most, written feedback before a score the least.	13.0 / 15
Match score	78.9 / 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 M2 Max (32 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q4_K_M`	4.7 GB	21.4 GB	331 ms	Runs well	72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	5.1 GB	21.4 GB	376 ms	Runs well	73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	5.5 GB	21.4 GB	425 ms	Slow	75Great quality (A) · 75 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q8_0`	6.4 GB	21.4 GB	536 ms	Slow	76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	9.9 GB	21.4 GB	1.0 s	Slow	76Great quality (A) · 76 of 100: how capable the model is at F16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab) Original weights (opens in a new tab)

Times are for one request with a 256-token input and about 24 tokens written. Batching servers process many requests at once.

Other hardware that runs it

76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5090 Q8_0 · 97 ms · Runs great
76Great quality (A) · 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA RTX PRO 6000 Blackwell Q8_0 · 95 ms · Runs great
73Great quality (A) · 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4090 Q5_K_M · 108 ms · Runs great
72Great quality (A) · 72 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 GeForce RTX 5080 Q4_K_M · 104 ms · Runs great
72Great quality (A) · 72 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. AMD Radeon RX 7900 XTX Q4_K_M · 105 ms · Runs great
72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5070 Ti Q4_K_M · 116 ms · Runs great

Quick answers

How much memory does NuExtract 2.0 4B need on the M2 Max 32GB?

At Q5_K_M it needs about 5.1 GB. The M2 Max 32GB has 21.4 GB usable.

How fast is NuExtract 2.0 4B on the M2 Max 32GB?

It makes one decision in about 376 ms at Q5_K_M, an estimate from the hardware and the size of the model.

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.