mlx-optiq
← Models

Qwen3.5-35B-A3B-OptiQ-4bit

Qwen/Qwen3.5-35B-A3B, quantized by OptiQ for Apple Silicon; mixture-of-experts, takes images.

visionMoEMTP head
On disk23.5 GB
Memoryabout 28 GB
Capability Score77.42
Imagesyes

Capability Score breakdown

The six metrics behind this model's Capability Score of 77.42, as reported in the model card. Values are compared with uniform 4-bit.

  • This model
  • Uniform 4-bit
Capability Score breakdown: Qwen3.5-35B-A3B-OptiQ-4bitMMLU: This model 86.7, Uniform 4-bit 85.9; GSM8K: This model 89.9, Uniform 4-bit 87.5; IFEval: This model 71.9, Uniform 4-bit 73.2; BFCL-V3: This model 93.0, Uniform 4-bit 89.5; HumanEval: This model 89.0, Uniform 4-bit 88.4; HashHop: This model 34.0, Uniform 4-bit 37.0MMLU86.785.9GSM8K89.987.5IFEval71.973.2BFCL-V393.089.5HumanEval89.088.4HashHop34.037.0

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 mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

OptiQ Code

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

optiq serve --model mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit
optiq code

OptiQ Lab

optiq lab --model mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

Claude Code

optiq run claude -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

Codex

optiq run codex -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

OpenCode

optiq run opencode -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

Mistral Vibe

optiq run vibe -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

Hermes Agent

optiq run hermes -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

OpenClaw

optiq run openclaw -m mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit

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("mlx-community/Qwen3.5-35B-A3B-OptiQ-4bit")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

FamilyQwen · Qwen docs
ArchitectureQwen3.5 MoE
Base modelQwen/Qwen3.5-35B-A3B
Size on disk23.5 GB
Memory neededabout 28 GB
Imagesyes, through optiq serve and the Lab
MTP headincluded
Capability Score77.42 (from the model card)
Downloads389 in the last month
Updated2026-07-14
Hugging Facemlx-community/Qwen3.5-35B-A3B-OptiQ-4bit