Ministral 3 3B on Intel Celeron J3455 / N3450 (Apollo Lake) NAS (16 GB)

Can the Intel Celeron J3455 / N3450 (Apollo Lake) NAS (16 GB) run Ministral 3 3B?

Barely: it runs very slowly

Very slow. At Q6_K it needs about 5.6 GB of 12.0 GB usable, with an estimated speed of 0.9 tok/s.

Best quant: Q6_K
Generation speed: 0.9 tok/s
Prompt processing: 1.6 tok/s
First token (1K prompt): 657.1 s
Largest safe context: 66K

Best app to run it on the Intel Celeron J3455 / N3450 (Apollo Lake) NAS (16 GB)

On Linux

Ollama

Command line + local API · runs on llama.cpp (ggml), MLX (Apple silicon, preview)

One command with Ollama’s own build of this model.
Serves an OpenAI-compatible API that other apps and coding tools can use.

ollama run ministral-3:3b

llama.cpp Usually the first to support new model architectures.
KoboldCpp

How we choose: the files published for this model, which apps run in a container on a NAS without a graphics card (models run on its processor), and whether the model fits fully in the memory models can use. Details.

Download Ministral 3 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 ministral-3:3b

Quantized files · GGUF

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

huggingface.co/mistralai/Ministral-3-3B-Instruct-2512-GGUF (opens in a new tab)

ollama run hf.co/mistralai/Ministral-3-3B-Instruct-2512-GGUF

Original weights · Hugging Face

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

huggingface.co/mistralai/Ministral-3-3B-Instruct-2512 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

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

github.com/mistralai/mistral-inference (opens in a new tab)

Mistral AI’s official inference code.

git clone https://github.com/mistralai/mistral-inference

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

How it’s rated

Quality 43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q6_K, 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 Very slow: How well it runs on the Intel Celeron J3455 / N3450 (Apollo Lake) NAS (16 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.	19.6 / 45
Memory fitFull marks when the model fits in the share of the NAS’s memory that models can use (the rest stays with the NAS system and its storage services); 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.	2.3 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	10.7 / 15
Match score	57.6 / 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 Intel Celeron J3455 / N3450 (Apollo Lake) NAS (16 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	4.0 GB	12.0 GB	2.2 tok/s	Very slow	30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q3_K_M`	4.3 GB	12.0 GB	1.7 tok/s	Very slow	36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q4_K_M`	4.8 GB	12.0 GB	1.3 tok/s	Very slow	41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	5.2 GB	12.0 GB	1.1 tok/s	Very slow	42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	5.6 GB	12.0 GB	0.9 tok/s	Very slow	43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q8_0`	6.5 GB	12.0 GB	0.7 tok/s	Very slow	44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	10.1 GB	12.0 GB	0.3 tok/s	Very slow	44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking.	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

44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4060 Q8_0 · 44.1 tok/s · Runs great
44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 9070 XT F16 · 41.2 tok/s · Runs great
44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4070 Ti SUPER F16 · 50.3 tok/s · Runs great
44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4080 SUPER F16 · 55.1 tok/s · Runs great
44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5070 Ti F16 · 67.0 tok/s · Runs great
44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5080 F16 · 71.8 tok/s · Runs great

Quick answers

How much memory does Ministral 3 3B need on the Intel Celeron J3455 / N3450 (Apollo Lake) NAS 16GB?

At Q6_K it needs about 5.6 GB. The Intel Celeron J3455 / N3450 (Apollo Lake) NAS 16GB has 12.0 GB usable.

How fast is Ministral 3 3B on the Intel Celeron J3455 / N3450 (Apollo Lake) NAS 16GB?

About 0.9 tok/s at Q6_K, an estimate from the hardware’s memory bandwidth and the size of the model.

Which quantization of Ministral 3 3B should I use on the Intel Celeron J3455 / N3450 (Apollo Lake) NAS 16GB?

Q6_K. Of the versions that fit, it is the one that runs best; when two run equally well, the check prefers the higher-rated one.

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.