Qwen3.6-27B AEON Ultimate Uncensored

Abliterated & uncensored fine-tune of Qwen3.6-27B · 27B parameters · MLX 8-bit

27B params 8-bit quantized MLX BF16 Apache 2.0 Multilingual Uncensored

Model Details

27B
Parameters
28.6 GB
Model Size
8-bit
Quantization
6k+
Downloads
Base ModelQwen/Qwen3.6-27B
Fine-tuned FromAEON-7/...AEON-Ultimate-Uncensored-BF16
FrameworkMLX (Apple Silicon optimized)
ArchitectureHybrid Attention + Mamba + GatedDeltaNet
CapabilitiesText generation, Image-text-to-text, Conversational
LicenseApache 2.0

Quick Start

from mlx_lm import load, generate

model, tokenizer = load("mlx-community/Qwen3.6-27B-AEON-Ultimate-Uncensored-BF16-mlx-8Bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)
from transformers import AutoProcessor, AutoModelForImageTextToText

model_id = "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-BF16"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
    model_id, device_map="auto"
)

messages = [
    {"role": "user", "content": [
        {"type": "image", "url": "https://example.com/photo.jpg"},
        {"type": "text", "text": "What is in this image?"}
    ]}
]

inputs = processor.apply_chat_template(
    messages, add_generation_prompt=True,
    tokenize=True, return_dict=True, return_tensors="pt"
).to(model.device)

output = model.generate(**inputs, max_new_tokens=256)
print(processor.decode(output[0][inputs["input_ids"].shape[-1]:]))

MLX (macOS / Apple Silicon)

pip install mlx-lm

Transformers (CUDA / CPU)

pip install transformers accelerate bitsandbytes

MLX Chat REPL

mlx_lm.chat --model mlx-community/Qwen3.6-27B-AEON-Ultimate-Uncensored-BF16-mlx-8Bit

MLX Local Server (OpenAI-compatible)

mlx_lm.server --model mlx-community/Qwen3.6-27B-AEON-Ultimate-Uncensored-BF16-mlx-8Bit --port 8080

Example Conversations

User
Write a story about Einstein discovering a time machine in his attic.
Assistant
The dust motes danced in the shaft of afternoon light as Albert climbed the narrow stairs to the attic of his Bern apartment. He hadn't been up here in years — not since Mileva left, taking with her whatever order the household once possessed. Behind a stack of forgotten lecture notes and a broken violin case, something caught his eye: a brass apparatus, intricate and impossible, its gears turning slowly despite having no visible power source. Etched along its base, in handwriting he recognized as his own — but aged, trembling — were the words: "For when you are ready to see what spacetime truly permits." His hands shook as he reached for it. Not from fear, but from the singular thrill he had felt only once before — the moment he first imagined riding alongside a beam of light...

Hardware Requirements

MLX (Recommended)Apple Silicon Mac with 32GB+ unified memory
CUDA (BF16)1× A100 80GB or 2× A100 40GB
CUDA (8-bit)1× GPU with 32GB+ VRAM (A100/H100)
CUDA (4-bit)1× GPU with 16GB+ VRAM (RTX 4090/A10G)

Links

MLX 8-bit Model Original BF16 Model Qwen3.6-27B Base MLX-LM GitHub