Qwen3.5-9b-DarkRoast-OptiQ-4bit
DavidAU/Qwen3.5-9B-The-Defiant-Fable-DARK-ROAST-Uncensored-Heretic-NEO-IMATRIX-MAX-MTP, quantized with OptiQ by jaithree for Apple Silicon; takes images.
communityvisionMTP headOn disk8.2 GB
Memoryabout 11 GB
Imagesyes
Use it
Install
pip install mlx-optiqServe it
An OpenAI- and Anthropic-compatible server on this Mac. Every app below can use it.
optiq serve --model jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitOptiQ Code
The terminal coding agent uses whatever model optiq serve is running.
optiq serve --model jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bit
optiq codeOptiQ Lab
optiq lab --model jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitClaude Code
optiq run claude -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitCodex
optiq run codex -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitOpenCode
optiq run opencode -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitMistral Vibe
optiq run vibe -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitHermes Agent
optiq run hermes -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitOpenClaw
optiq run openclaw -m jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bitoptiq 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("jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bit")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))Details
| Family | Qwen · Qwen docs |
|---|---|
| Architecture | Qwen3.5 |
| Base model | DavidAU/Qwen3.5-9B-The-Defiant-Fable-DARK-ROAST-Uncensored-Heretic-NEO-IMATRIX-MAX-MTP |
| Size on disk | 8.2 GB |
| Memory needed | about 11 GB |
| Images | yes, through optiq serve and the Lab |
| MTP head | included |
| Downloads | 160 in the last month |
| Updated | 2026-08-07 |
| Hugging Face | jaithree/Qwen3.5-9b-DarkRoast-OptiQ-4bit |