mlx-optiq
← Models

Qwen3.6-27B-4.2Bit-OptiQ-MTP

Qwen/Qwen3.6-27B, quantized with OptiQ by iamthecage for Apple Silicon.

community
On disk17.3 GB
Memoryabout 21 GB

Use it

Install

pip install mlx-optiq

Serve it

An OpenAI- and Anthropic-compatible server on this Mac. Every app below can use it.

optiq serve --model iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

OptiQ Code

The terminal coding agent uses whatever model optiq serve is running.

optiq serve --model iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP
optiq code

OptiQ Lab

optiq lab --model iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

Claude Code

optiq run claude -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

Codex

optiq run codex -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

OpenCode

optiq run opencode -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

Mistral Vibe

optiq run vibe -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

Hermes Agent

optiq run hermes -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

OpenClaw

optiq run openclaw -m iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP

optiq run starts optiq serve if it is not running, sets the agent up for this session only, and stops the server when the agent exits.

mlx-lm, text only

from mlx_lm import load, generate

model, tokenizer = load("iamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

FamilyQwen · Qwen docs
ArchitectureQwen3.5
Base modelQwen/Qwen3.6-27B
Size on disk17.3 GB
Memory neededabout 21 GB
Imagesno, text only
Downloads28 in the last month
Updated2026-06-24
Hugging Faceiamthecage/Qwen3.6-27B-4.2Bit-OptiQ-MTP