Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-mxfp4-mlx

Brainwaves

         arc   arc/e boolq hswag obkqa piqa  wino
mxfp8   0.665,0.831,0.910,0.790,0.456,0.813,0.772

Quant    Perplexity      Peak Memory   Tokens/sec
mxfp8    5.073 ± 0.038   34.74 GB      212
mxfp4    4.950 ± 0.036   21.30 GB      187

Baseline model

         arc   arc/e boolq hswag obkqa piqa  wino
mxfp8    0.647,0.803,0.910,0.773,0.450,0.806,0.742
qx86-hi  0.637,0.798,0.911,0.775,0.442,0.807,0.737

This model is using the fixed jinja template from froggeric/Qwen-Fixed-Chat-Templates

Thinking toggle

Drop <|think_on|> or <|think_off|> anywhere in your system or user prompt. The template intercepts the tag, removes it from context so the model never sees it, and flips the mode.

Fast answer, no reasoning:

System: You are a coding assistant. <|think_off|>
User: What's 2+2?

Deep reasoning:

System: You are a coding assistant. <|think_on|>
User: Implement a red-black tree in Rust.

The tag syntax (<|think_on|>, <|think_off|>) uses Qwen's control-token delimiters, so it will never collide with real text. Earlier community templates used /think, which broke legitimate paths like cd /mnt/project/think.

---

I added a similar set of tags for handling the preserve_thinking flag:

• Drop <|think_forget|> or <|think_remember|> anywhere in your system or user prompt to flip the flag.
• The template intercepts the tag, removes it from context so the model never sees it, and flips the mode.

-G

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-mxfp4-mlx")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True, return_dict=False,
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)