LLM EXPLORER 65,418 MODELS INDEXED

Jugurtha 4B MLX 4bit by agbalu

By agbalu · 157 downloads
HF Repo ↗ Copy URL

Jugurtha 4B MLX 4bit is an open-source language model by agbalu. Features: 4b LLM, VRAM: 2.4GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.31.

  4-bit   4bit   Amazigh   Apple-silicon   Base model:agbalu/jugurtha-4b Base model:quantized:agbalu/ju...   Berber   Conversational   En   Fr   Kab   Kabyle   Mlx   Mlx-my-repo   Quantized   Qwen3 5   Reasoning   Region:us   Safetensors   Sovereign-ai   Taqbaylit   Thinking
Model Card on HF 🤗: https://huggingface.co/agbalu/Jugurtha-4B-MLX-4bit 

Jugurtha 4B MLX 4bit Parameters and Internals

LLM NameJugurtha 4B MLX 4bit
Repository 🤗https://huggingface.co/agbalu/Jugurtha-4B-MLX-4bit 
Base Model(s)  Jugurtha 4B   agbalu/Jugurtha-4B
Model Size4b
Required VRAM2.4 GB
Updated2026-10-22
Maintaineragbalu
Model Typeqwen3_5
Model Files  2.4 GB
Supported Languagesen fr
Quantization Type4bit
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.0.dev0
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Jugurtha 4B MLX 4bit

Best Alternatives
Context / RAM
Downloads
Likes
Underdog Woof 4B 1.10K / 2.4 GB186623
H2o Lightning 4B Mlx Q80K / 5.1 GB1582
Qwen3.5 4B MLX 4bit0K / 3 GB2802332
LightOnOCR 3 4B 4bit0K / 3 GB1771
...Bonsai 2 4B ONNX WebGPU Vision0K /  GB691
Loopl 4B 4bit0K / 3 GB1701
H2o Lightning 4B Mlx Q40K / 3 GB951
Sbzhanzhang0K / 2.4 GB1581
Qwen3.5 4B OptiQ 4bit0K / 0.1 GB782323
Qwen3.5 4B 4bit Abliterated0K / 3 GB851
Note: green Score (e.g. "73.2") means that the model is better than agbalu/Jugurtha-4B-MLX-4bit.