Qwen3.8-27B-Uncensored-OptiQ-4bit

mlx-community/Qwen3.8-27B-Uncensored-OptiQ-4bit

Built with mlx-optiq, the MLX-native toolkit to quantize, fine-tune, and serve LLMs locally on Apple Silicon, no PyTorch and no cloud. All OptiQ quants · Docs

OptiQ mixed-precision quant of orcarouter/Qwen3.8-27B-Uncensored, a Qwen3.8-family reasoning model with a bundled MTP speculation head. 19 GB on disk.

What it is

Following the naming llama.cpp uses for its mixed quants, the "4bit" label denotes the family, not the weighted average.

Run it

Qwen3.8 and the MTP sidecar register through OptiQ, so import optiq once before loading:

pip install "mlx-optiq>=0.4.27"

import optiq  # registers the arch + MTP sidecar
from mlx_lm import load, generate

model, tok = load("mlx-community/Qwen3.8-27B-Uncensored-OptiQ-4bit")
prompt = tok.apply_chat_template(
    [{"role": "user", "content": "Explain mixed-precision quantization in two sentences."}],
    tokenize=False, add_generation_prompt=True,
)
print(generate(model, tok, prompt=prompt, max_tokens=400))

For an OpenAI- and Anthropic-compatible endpoint with mixed-precision KV cache:

optiq serve --model mlx-community/Qwen3.8-27B-Uncensored-OptiQ-4bit

This is a reasoning model, so give it a generous token budget.

Using the head outside optiq serve: optiq/mtp.safetensors holds Qwen's original MTP tensors, so its seven RMSNorm weights (mtp.*norm*) are zero-centred, applied as 1 + w. The main weights have the +1 folded in, as MLX expects. Add 1.0 to all seven norms when loading the head into another runtime.

Links

• Project website: mlx-optiq.com
• All OptiQ quants: mlx-optiq.com/models
• Base model: orcarouter/Qwen3.8-27B-Uncensored