Can the Apple M2 (24 GB) run d1 3B?

Yes, it runs well

Runs well. At Q5_K_M it needs about 4.1 GB of 16.1 GB usable and makes one decision in about 398 ms.

Best quant: Q5_K_M
Time per decision: 398 ms
Decisions per second: 2.5/s
Input assumed: 256 tokens
Returns: Probability or score

Best app to run it on the Apple M2 (24 GB)

On macOS

Python library · runs on PyTorch (Transformers)

A few lines of Python with Transformers or Sentence Transformers; easiest for scripts and notebooks.

python -c "from transformers import pipeline; print(pipeline('text-classification', model='LiquidAI/d1-3B')('text'))"

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 d1 3B

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 d1:3b

Quantized files · GGUF

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

huggingface.co/LiquidAI/d1-3B-GGUF (opens in a new tab)

ollama run hf.co/LiquidAI/d1-3B-GGUF

Original weights · Hugging Face

Full-precision files from Liquid AI, for Transformers, Sentence Transformers, Text Embeddings Inference, Infinity or vLLM.

huggingface.co/LiquidAI/d1-3B (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 79Great quality (A) · 79 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (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 (24 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.	35.8 / 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.	7.5 / 15
ClarityHow easy the answer is to use in code: a probability or score earns the most, written feedback before a score the least.	15.0 / 15
Match score	83.3 / 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 (24 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q4_K_M`	3.8 GB	16.1 GB	391 ms	Runs well	77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	4.1 GB	16.1 GB	398 ms	Runs well	79Great quality (A) · 79 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q6_K`	4.5 GB	16.1 GB	404 ms	Slow	81Great quality (A) · 81 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q8_0`	5.2 GB	16.1 GB	420 ms	Slow	82Great quality (A) · 82 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	8.2 GB	16.1 GB	488 ms	Slow	82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (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 1 tokens written. Batching servers process many requests at once.

Other hardware that runs it

82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 7900 XTX F16 · 28 ms · Runs great
82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 9070 XT F16 · 37 ms · Runs great
82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. AMD Ryzen AI Max+ 395 (128 GB) F16 · 75 ms · Runs great
82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. Apple M2 Ultra (192 GB) F16 · 67 ms · Runs great
82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. Apple M3 Ultra (256 GB) F16 · 65 ms · Runs great
82Great quality (A) · 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. Apple M3 Ultra (512 GB) F16 · 65 ms · Runs great

Quick answers

How much memory does d1 3B need on the M2 24GB?

At Q5_K_M it needs about 4.1 GB. The M2 24GB has 16.1 GB usable.

How fast is d1 3B on the M2 24GB?

It makes one decision in about 398 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.