Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

Qwen/Qwen-AgentWorld-35B-A3B, quantized with OptiQ by Aminlight for Apple Silicon; mixture-of-experts.

communityMoE

On disk36.3 GB

Memoryabout 43 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 Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

OptiQ Code

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

optiq serve --model Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX
optiq code

OptiQ Lab

optiq lab --model Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

Claude Code

optiq run claude -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

Codex

optiq run codex -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

OpenCode

optiq run opencode -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

Mistral Vibe

optiq run vibe -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

Hermes Agent

optiq run hermes -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

OpenClaw

optiq run openclaw -m Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX

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("Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX")
print(generate(model, tokenizer, prompt="Hello", max_tokens=256))

Details

Family	Qwen · Qwen docs
Architecture	Qwen3.5 MoE
Base model	Qwen/Qwen-AgentWorld-35B-A3B
Size on disk	36.3 GB
Memory needed	about 43 GB
Images	no, text only
Downloads	29 in the last month
Updated	2026-06-24
Hugging Face	Aminlight/Qwen-AgentWorld-35B-A3B-OptiQ-6bpw-MLX