Skip to content

Ministral-3-14B-Base-2512-6bit

mlx-community/Ministral-3-14B-Base-2512-6bit
6-bit Open on Hugging Face
Parameters
14B
Weights (est.)
≈ 12.6 GB
Mac memory (est.)
24 GB+

Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.

Run it

With mlx-vlm on an Apple silicon Mac:

pip install mlx-vlmpython -m mlx_vlm.generate --model mlx-community/Ministral-3-14B-Base-2512-6bit --image photo.jpg --prompt "Describe this image"

Details

Base model
mistralai/Ministral-3-14B-Base-2512 (quantized)
Vendor
Mistral
Task
Vision + text
Quantisation
6-bit
Licence
apache-2.0
Downloads
85 in the last 30 days · 323 all time · 0 likes
Published
2026-07-18 · updated 2026-07-18 (3mo ago)
Languages
en, fr, es, de, it, pt, nl, zh, ja, ko, ar
Tags
mistral3, mistral-common, ministral, ministral-3, vision-language, multimodal, quantized, edge, base-model

Other MLX builds of mistralai/Ministral-3-14B-Base-2512

Build Quantisation Weights Mac memory Downloads
Ministral-3-14B-Instruct-2512-mxfp4 MXFP4 ≈ 8.4 GB 16 GB+ 85
Ministral-3-14B-Instruct-2512-nvfp4 4-bit ≈ 8.4 GB 16 GB+ 41
Ministral-3-14B-Base-2512-4bit 4-bit ≈ 9.4 GB 16 GB+ 125
Ministral-3-14B-Base-2512-6bit (this one) 6-bit ≈ 12.6 GB 24 GB+ 85
Ministral-3-14B-Instruct-2512-mxfp8 MXFP8 ≈ 15.2 GB 24 GB+ 65
Ministral-3-14B-Base-2512-8bit 8-bit ≈ 15.9 GB 24 GB+ 101
Ministral-3-14B-Reasoning-2512-bf16 bf16 ≈ 27.9 GB 48 GB+ 134
Ministral-3-14B-Base-2512-bf16 bf16 ≈ 27.9 GB 48 GB+ 82
Ministral-3-14B-Instruct-2512-4bit 4-bit — — 3.5K
Ministral-3-14B-Reasoning-2512-4bit 4-bit — — 786
Ministral-3-14B-Reasoning-2512-8bit 8-bit — — 741
Ministral-3-14B-Instruct-2512-8bit 8-bit — — 350
Ministral-3-14B-Instruct-2512-6bit 6-bit — — 113
Ministral-3-14B-Instruct-2512-bf16 bf16 — — 75
Ministral-3-14B-Reasoning-2512-6bit 6-bit — — 75
Ministral-3-14B-Reasoning-2512-5bit 5-bit — — 56
Ministral-3-14B-Instruct-2512-5bit 5-bit — — 37