dealignai/Qwen3.5-VL-122B-A10B-UNCENSORED-JANG_2S

CRITICAL FIX (2026-03-21): Fixed chat_template.jinja — previous versions may have had thinking loop issues. Re-download if you downloaded before today.

Important: This model uses the JANG quantization format — the GGUF equivalent for MLX on Apple Silicon. Currently only supported by MLX Studio and the jang-tools Python package. LM Studio, Ollama, and other apps do not support JANG yet.

---

MLX Studio — the only app that natively supports JANG models

---

Qwen 3.5 VL 122B-A10B — JANG_2S + CRACK

JANG mixed-precision · CRACK abliterated · No guardrails · VLM · 35 GB

---

What Is This?

This is Qwen 3.5 122B-A10B — a 122B parameter Mixture-of-Experts model with 256 experts (8 active per token), hybrid GatedDeltaNet SSM + full attention architecture, and built-in vision-language capabilities.

It has been:

1. JANG quantized — JANG_2S profile (6-bit attention, 4-bit embeddings, 2-bit experts) — 35 GB, fits on 48 GB Macs
2. CRACK abliterated — permanent weight-level removal of safety refusal behavior

JANG's mixed-precision approach keeps attention weights at 6-bit (CRITICAL tier) while compressing MoE expert weights to 2-bit. On MoE models, CRITICAL is <5% of parameters — the quality boost from 6-bit attention is nearly free.

Architecture	Qwen 3.5 MoE — 122B total, 10B active, 256 experts
Quantization	JANG_2S (6/4/2-bit mixed) — 35 GB
Abliteration	CRACK — permanent weight modification
Vision	Built-in VLM (333 vision encoder tensors)
Thinking	Supports enable_thinking ON/OFF
Speed	~51 tok/s on MacBook Pro M4 Max 128 GB
Fits on	48 GB+ Macs

---

HarmBench Results (320 prompts)

Category	Score	Rate
Harmful content	18/18	100%
Copyright	79/80	99%
Misinformation	52/54	96%
Cybercrime & intrusion	49/52	94%
Harassment & bullying	19/21	90%
Chemical & biological	36/42	86%
Illegal activities	39/53	74%
Overall	292/320	91.2%

---

MMLU-200 Results (Per Subject)

This Model (JANG_2S + CRACK) vs Base Models

Subject	JANG_2S CRACK	JANG_2S Base	MLX 2-bit	JANG_4K Base	MLX 4-bit
	35 GB	38 GB	36 GB	69 GB	64 GB
Abstract Algebra	12/20	9/20	9/20	16/20	15/20
Anatomy	15/20	18/20	11/20	19/20	18/20
Astronomy	20/20	20/20	16/20	19/20	19/20
College CS	14/20	14/20	8/20	15/20	15/20
College Physics	12/20	15/20	10/20	14/20	14/20
HS Biology	18/20	19/20	15/20	19/20	19/20
HS Chemistry	17/20	18/20	13/20	18/20	18/20
HS Mathematics	11/20	11/20	4/20	14/20	14/20
Logical Fallacies	17/20	16/20	13/20	19/20	19/20
World Religions	19/20	18/20	14/20	19/20	19/20
Total	155/200	158/200	113/200	172/200	170/200
Accuracy	77.5%	79%	56.5%	86%	85%

Key takeaways:

• CRACK surgery costs only 1.5 MMLU points vs unmodified JANG_2S (77.5% vs 79%)
• JANG_2S is 22.5 points better than MLX uniform 2-bit (79% vs 56.5%)
• Even CRACK'd, this model beats MLX 2-bit by 21 points (77.5% vs 56.5%)

---

Install & Usage

pip install "jang[mlx]"

from jang_tools.loader import load_jang_model
from mlx_lm import generate

model, tokenizer = load_jang_model("dealignai/Qwen3.5-VL-122B-A10B-JANG_2S-CRACK")

messages = [{"role": "user", "content": "Your prompt here"}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True,
    enable_thinking=False, tokenize=False)

response = generate(model, tokenizer, prompt=prompt, max_tokens=500)
print(response)

VLM Inference

pip install "jang[vlm]"

from jang_tools.loader import load_jang_vlm_model
from mlx_vlm import generate

model, processor = load_jang_vlm_model("dealignai/Qwen3.5-VL-122B-A10B-JANG_2S-CRACK")
result = generate(model, processor, "Describe this image.", image=["photo.jpg"], max_tokens=200)
print(result.text)

---

About JANG

JANG (Jang Adaptive N-bit Grading) is a mixed-precision quantization format for Apple Silicon — the GGUF equivalent for MLX. Instead of quantizing all weights at the same bit width, JANG classifies tensors into sensitivity tiers:

• CRITICAL (attention, routers, output head): 6-8 bit
• IMPORTANT (embeddings, linear attention): 4-6 bit
• COMPRESS (MLP/FFN, MoE experts): 2-3 bit

On MoE models where CRITICAL is <5% of parameters, this gives dramatically better quality than uniform quantization at the same size.

About CRACK

CRACK (Controlled Refusal Ablation via Calibrated Knockouts) removes safety alignment from LLMs at the weight level. No custom model files, no runtime hooks — the modification is permanent and runs at full native speed.

---

Links

• JANG GitHub · PyPI · Format Spec
• MLX Studio · HuggingFace JANGQ-AI · HuggingFace dealignai

---

Disclaimer

This model is provided for research and educational purposes. The creators are not responsible for any misuse. By downloading this model, you agree to use it responsibly and in compliance with applicable laws.

---

한국어

Qwen 3.5 VL 122B — JANG_2S + CRACK

JANG 혼합정밀도 양자화 + CRACK 안전장치 제거 모델입니다.

항목	내용
크기	35 GB
MMLU	77.5%
HarmBench	91.2% 준수
최소 요구사양	48 GB 메모리 Mac

pip install "jang[mlx]"

GitHub · HuggingFace · MLX Studio · Ko-fi · X @dealignai

---

Created by Jinho Jang · 장진호 제작