          arc   arc/e boolq hswag obkqa piqa  wino
qx86-hi   0.443,0.498,0.857,0.701,0.372,0.770,0.752
qx64-hi   0.435,0.494,...
mxfp4     0.460,0.527,0.871,0.694,0.370,0.772,0.752

Quant     Perplexity      Peak Memory
qx86-hi   3.888 ± 0.025   33.34 GB
qx64-hi   3.917 ± 0.025   26.62 GB

Qwen3.5-27B-Text-heretic
qx64-hi   3.931 ± 0.025   26.62 GB

Similar thinking models

brainwaves

TeichAI/Qwen3.5-27b-Opus-4.6-Distill
qx64-hi   0.459,0.542,0.724,0.764,0.402,0.790,0.783

DavidAU/Qwen3.5-27B-Polaris-Advanced-Thinking-Alpha
mxfp4     0.473,0.548,0.709,0.728,0.396,0.777,0.753

DavidAU/Qwen3.5-27B-Claude-4.6-OS-Auto-Variable-Thinking
mxfp8     0.485,0.566,0.875,0.746,0.408,0.789,0.730

Instruct models

brainwaves

DavidAU/Qwen3.5-27B-Claude-4.6-OS-INSTRUCT
mxfp8     0.675,0.827,0.900,0.750,0.496,0.800,0.721
qx86-hi   0.667,0.822,0.900
qx64-hi   0.664,0.820,0.902
mxfp4     0.653,0.815,0.899

Older VL models

brainwaves

Huihui-Qwen3-VL-32B-Thinking-abliterated
qx86-hi   0.376,0.449,0.823,0.637,0.378,0.772,0.681

unsloth-Qwen3-VL-32B-Instruct
qx86x-hi  0.447,0.593,0.904,0.610,0.432,0.738,0.594

Specialized older VL models

brainwaves

nightmedia/Qwen3-32B-Element5-Heretic
qx86-hi   0.483,0.596,0.738,0.754,0.394,0.802,0.710

UIGEN-FX-Agentic-32B
qx86-hi   0.460,0.527,0.681,0.721,0.404,0.790,0.728

More metrics are compiled and will be available soon.

-G

This model Qwen3.5-27B-Text was converted to MLX format from Qwen/Qwen3.5-27B using mlx-lm version 0.30.8.

Use with mlx

pip install mlx-lm

python

from mlx_lm import load, generate

model, tokenizer = load("Qwen3.5-27B-Text")

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)

nightmedia/Qwen3.5-27B-Text

Qwen3.5-27B-Text

Use with mlx

nightmedia/
Qwen3.5-27B-Text