LLM EXPLORER 67,808 MODELS INDEXED

Grux Local 4B by gruxai

By gruxai · 24 downloads

Grux Local 4B is an open-source language model by gruxai. Features: 4b LLM, VRAM: 3.7GB, License: apache-2.0, LLM Explorer Score: 0.29.

  4-bit   Agent   Apple-silicon Base model:finetune:qwen/qwen3...   Base model:qwen/qwen3.5-4b   Conversational   En   Function-calling   Grux   Macos   Mlx   Model-index   Nvfp4   Prompt-injection   Qwen3 5   Region:us   Safetensors   Sharded   Tensorflow   Tool-calling
Model Card on HF πŸ€—: https://huggingface.co/gruxai/grux-local-4b 

Grux Local 4B Parameters and Internals

LLM NameGrux Local 4B
Repository πŸ€—https://huggingface.co/gruxai/grux-local-4b 
Base Model(s)  Qwen/Qwen3.5-4B   Qwen/Qwen3.5-4B
Model Size4b
Required VRAM3.7 GB
Updated2026-11-18
Maintainergruxai
Model Typeqwen3_5
Model Files  3.7 GB: 1-of-1
Supported Languagesen
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Grux Local 4B

Best Alternatives
Context / RAM
Downloads
Likes
AgentHorse 4B256K / 8.4 GB9737
Qwen3.5 4B0K / 9.3 GB6575502801
KORMo Web 4B0K / 10.4 GB21251
NuExtract30K / 0.2 GB441570324
Qwen3.5 4B MLX0K / 3 GB272281
Agents A1 4B0K / 9.1 GB17757872
LightOnOCR 3 4B0K / 9.1 GB239571
...en3.5 4B 3D Coder MTP Repacked0K / 9.3 GB21328
FrogNano 4B 26090K / 9.3 GB733113
Rein NPC 4B0K / 10.6 GB13663
Note: green Score (e.g. "73.2") means that the model is better than gruxai/grux-local-4b.