LLM EXPLORER 58,873 MODELS INDEXED

Ministral 3 8B Instruct 2512 by mistralai

By mistralai · 123761 downloads
HF Repo ↗ Copy URL

Ministral 3 8B Instruct 2512 is an open-source language model by mistralai. Features: 8b LLM, VRAM: 10.4GB, Context: 256K, License: MNPL-0.1, Instruction-Based, LLM Explorer Score: 0.31.

  Arxiv:2601.08584   Ar Base model:mistralai/ministral... Base model:quantized:mistralai...   De   En   Es   Fp8   Fr   Instruct   It   Ja   Ko   Mistral-common   Mistral3   Nl   Pt   Region:us   Safetensors   Sharded   Tensorflow   Vllm   Zh
Model Card on HF 🤗: https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512 

Ministral 3 8B Instruct 2512 Parameters and Internals

LLM NameMinistral 3 8B Instruct 2512
Repository 🤗https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512 
Base Model(s)  Ministral 3 8B Base 2512   mistralai/Ministral-3-8B-Base-2512
Model Size8b
Required VRAM10.4 GB
Updated2026-07-29
Maintainermistralai
Instruction-BasedYes
Model Files  10.4 GB   5.0 GB: 1-of-3   5.0 GB: 2-of-3   0.4 GB: 3-of-3
Supported Languagesen fr es de it pt nl zh ja ko ar
Model ArchitectureMambaSSM
LicenseMNPL-0.1
Context Length262144
Model Max Length262144
Tokenizer ClassTokenizersBackend
Padding Token<pad>
Vocabulary Size131072

Quantized Models of the Ministral 3 8B Instruct 2512

Model
Likes
Downloads
VRAM
Amaretto 8B137017 GB
...Instruct 2512 Unsloth Bnb 4bit129847 GB
...al 3 8B Instruct 2512 Bnb 4bit02556 GB

Best Alternatives to Ministral 3 8B Instruct 2512

Best Alternatives
Context / RAM
Downloads
Likes
...istral 3 8B Instruct 2512 BF16256K / 17.8 GB1401923
...8B Instruct 2512 BF16 SOM MPOA256K / 17.8 GB160
Ministral 8B Instruct 24100K / 16.1 GB299686584
Ministral 8B Instruct 2410 HF0K / 16.1 GB70
Note: green Score (e.g. "73.2") means that the model is better than mistralai/Ministral-3-8B-Instruct-2512.