LLM EXPLORER 61,561 MODELS INDEXED

ReAligned Qwen3.5 4B by Lazarus-Ai

By Lazarus-Ai · 126 downloads
HF Repo ↗ Copy URL

ReAligned Qwen3.5 4B is an open-source language model by Lazarus-Ai. Features: 4b LLM, VRAM: 9.1GB, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.23.

Base model:adapter:qwen/qwen3....   Base model:qwen/qwen3.5-4b   Censorship-mitigation   Chat   China-bias   Conspiracy   Conversational   Coverup   En   Endpoints compatible   Eric-hartford   Gaslighting   Grpo   Ideological-bias   Image-text-to-text   Instruct   Instruction-tuned   Lazarusai   Lora   Propaganda   Quixiai   Qwen3.5   Qwen3 5   Realigned   Region:us   Safetensors   Sft   Sharded   Tensorflow   Zh
Model Card on HF 🤗: https://huggingface.co/Lazarus-Ai/ReAligned-Qwen3.5-4B 

ReAligned Qwen3.5 4B Parameters and Internals

LLM NameReAligned Qwen3.5 4B
Repository 🤗https://huggingface.co/Lazarus-Ai/ReAligned-Qwen3.5-4B 
Base Model(s)  Qwen/Qwen3.5-4B   Qwen/Qwen3.5-4B
Model Size4b
Required VRAM9.1 GB
Updated2026-08-16
MaintainerLazarus-Ai
Model Typeqwen3_5
Instruction-BasedYes
Model Files  5.0 GB: 1-of-2   4.1 GB: 2-of-2
Supported Languagesen zh
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to ReAligned Qwen3.5 4B

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.5 4B VerIH Step2000K / 18.2 GB1920
Note: green Score (e.g. "73.2") means that the model is better than Lazarus-Ai/ReAligned-Qwen3.5-4B.