LLM EXPLORER 66,143 MODELS INDEXED

Qwen3.5 4B Abliterated by cfahlgren1

By cfahlgren1 · 231 downloads
HF Repo ↗ Copy URL

Qwen3.5 4B Abliterated is an open-source language model by cfahlgren1. Features: 4b LLM, VRAM: 9.3GB, License: apache-2.0, LLM Explorer Score: 0.31.

  Arxiv:2406.11717   Abliterated Base model:finetune:qwen/qwen3...   Base model:qwen/qwen3.5-4b   Conversational Dataset:mlabonne/harmful behav... Dataset:mlabonne/harmless alpa...   Endpoints compatible   Image-text-to-text   Mlx   Qwen3 5   Region:us   Safetensors   Sharded   Tensorflow   Uncensored
Model Card on HF 🤗: https://huggingface.co/cfahlgren1/Qwen3.5-4B-abliterated 

Qwen3.5 4B Abliterated Parameters and Internals

LLM NameQwen3.5 4B Abliterated
Repository 🤗https://huggingface.co/cfahlgren1/Qwen3.5-4B-abliterated 
Base Model(s)  Qwen/Qwen3.5-4B   Qwen/Qwen3.5-4B
Model Size4b
Required VRAM9.3 GB
Updated2026-10-30
Maintainercfahlgren1
Model Typeqwen3_5
Model Files  5.3 GB: 1-of-2   4.0 GB: 2-of-2
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Qwen3.5 4B Abliterated

Best Alternatives
Context / RAM
Downloads
Likes
AgentHorse 4B256K / 8.4 GB9737
Qwen3.5 4B0K / 9.3 GB6575502801
NuExtract30K / 0.2 GB441570324
LightOnOCR 3 4B0K / 9.1 GB239571
Agents A1 4B0K / 9.1 GB17757872
FrogNano 4B 26090K / 9.3 GB733113
UserRepGen 4B0K / 9.1 GB100813
Qwen3.5 4B Base0K / 9.3 GB22781479
Intern Decision 4B0K / 0.1 GB33848
Kestrel 4B Coder0K / 9.3 GB2513
Note: green Score (e.g. "73.2") means that the model is better than cfahlgren1/Qwen3.5-4B-abliterated.