Skip to content

Text model · Mistral

Ministral 3 14B: RAM and VRAM requirements

Ministral 3 14B needs about 9.4 GB to run at Q4_K_M (the Q4_K_M GGUF file is ~7.67 GB to download; KV cache and overhead add the rest), or about 15.1 GB at Q8_0. The lightest hardware that runs it is Nvidia GeForce RTX 3060 (12GB).

Mistral family · 14B params · released Dec 2025 · 1.4M Ollama pulls.

Shopping for hardware? See what runs Ministral 3 14B →

License Apache-2.0 · Commercial OK ↓ 327.7K/mo ♥ 320 on HuggingFace
Q4_K_M GGUF
7.67 GB
Q8_0 GGUF
13.37 GB
Memory @ Q4 (4k)
~9.4 GB
Context
256 k

Q4_K_M and Q8_0 are GGUF download (file) sizes. Memory @ Q4 includes the KV cache and overhead, so it runs larger than the file on disk.

Will it run on your device?

Ministral 3 14B runs on 29 of 43 tracked devices at Q4_K_M.

29 run well 0 tight fit 14 too small
Fit check Q4_K_M
Yes, it runs fast
needs 9.4 GB usable 10.5 GB
See the full breakdown
$ollama run ministral-3:14b

Memory breakdown

Weights (Q4_K_M)7.67 GB
+
KV cache (4k)0.9 GB
+
Overhead0.8 GB
=
Total9.4 GB

How context length changes it

4k context ~9.4 GB 32k context ~15.7 GB 128k context ~37.2 GB

Longer context grows the KV cache quickly: Ministral 3 14B needs ~9.4 GB at 4k but ~37.2 GB at 128k, which can push it past a device that fits it at a short context.

Speed drops too: every token re-reads the KV cache, so at 128k context Ministral 3 14B generates at roughly ~21% of its short-context speed. How this is estimated.

Quantization sizes

GGUF quantson disk
Quantization Size on disk
Q2_K 5.9 GB est
Q3_K_M 6.8 GB est
Q4_K_M (default) 7.67 GB
Q5_K_M 10 GB est
Q6_K 11.5 GB est
Q8_0 13.37 GB
FP16 27.02 GB

Lower quant = smaller and faster, slightly lower quality. Q4_K_M is the common default.

Run it

Ollama
$ ollama run ministral-3:14b
llama.cpp
$ llama-cli -hf mistralai/Ministral-3-14B-Instruct-2512-GGUF:Q4_K_M
LM Studio
$ lms get mistralai/Ministral-3-14B-Instruct-2512-GGUF

Which devices can run Ministral 3 14B?

Same job, different size

Pick by what fits your memory: step down to free up VRAM, or step up if you have headroom.

Similar models

Head-to-head

FAQ

How much VRAM or RAM does Ministral 3 14B need?

At Q4_K_M, Ministral 3 14B needs about 9.4 GB (weights ~7.67 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~15.1 GB.

What is the Q4_K_M GGUF file size of Ministral 3 14B?

The Q4_K_M GGUF file is about 7.67 GB to download, and the Q8_0 GGUF is about 13.37 GB. That is the weights file on disk; to run it you also need room for the KV cache and overhead, so budget ~9.4 GB of memory at Q4_K_M.

Can Ministral 3 14B run on a laptop?

Ministral 3 14B is large; you need a 24 GB+ GPU or a 32-48 GB Mac at Q4_K_M.

Can I use Ministral 3 14B commercially?

Yes. Ministral 3 14B is licensed Apache-2.0, which permits commercial use.

Understand the numbers

Short guides to the ideas behind Ministral 3 14B's memory and quant figures.

Largest of Mistral's edge family: dense 14B with vision, 256K context, Apache-2.0. Sizes from Mistral's own GGUF repo (the Reasoning variant ships at identical sizes).

Sources

Last validated 2026-09-07. Memory figures are estimates. See methodology.