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.
1from jang_tools.loader import load_jang_model
2from mlx_lm import generate
3
4model, tokenizer = load_jang_model("dealignai/Qwen3.5-VL-122B-A10B-JANG_2S-CRACK")
5
6messages = [{"role": "user", "content": "Your prompt here"}]
7prompt = tokenizer.apply_chat_template(
8 messages, add_generation_prompt=True,
9 enable_thinking=False, tokenize=False)
10
11response = generate(model, tokenizer, prompt=prompt, max_tokens=500)
12print(response)
1from jang_tools.loader import load_jang_vlm_model
2from mlx_vlm import generate
3
4model, processor = load_jang_vlm_model("dealignai/Qwen3.5-VL-122B-A10B-JANG_2S-CRACK")
5result = generate(model, processor, "Describe this image.", image=["photo.jpg"], max_tokens=200)
6print(result.text)
On MoE models where CRITICAL is <5% of parameters, this gives dramatically better quality than uniform quantization at the same size.
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.
