Instructions to use AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4 with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4")
config = load_config("AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4

Run Hermes

hermes

OpenClaw new

How to use AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4 with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "AEON-7/Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Qwen3.6-27B-AEON-Ultimate-Uncensored-Multimodal-MLX-FP4

File size: 2,852 Bytes

da35640

<svg viewBox="0 0 760 440" xmlns="http://www.w3.org/2000/svg" font-family="system-ui,-apple-system,Segoe UI,Roboto,sans-serif"><rect width="760" height="440" rx="14" fill="#0d1117"/><rect x="1" y="1" width="758" height="438" rx="13" fill="none" stroke="#30363d"/><text x="28" y="42" fill="#e6edf3" font-size="22" font-weight="700">Pick your build — speed vs footprint</text><text x="28" y="66" fill="#8b949e" font-size="13">bubble = effective precision (bits/weight) · top-left wins · M4 Pro 48 GB</text><line x1="80" y1="110" x2="680" y2="110" stroke="#30363d"/><text x="72" y="114" fill="#8b949e" font-size="10" text-anchor="end">30</text><line x1="80" y1="172" x2="680" y2="172" stroke="#30363d"/><text x="72" y="176" fill="#8b949e" font-size="10" text-anchor="end">22</text><line x1="80" y1="235" x2="680" y2="235" stroke="#30363d"/><text x="72" y="239" fill="#8b949e" font-size="10" text-anchor="end">15</text><line x1="80" y1="298" x2="680" y2="298" stroke="#30363d"/><text x="72" y="302" fill="#8b949e" font-size="10" text-anchor="end">8</text><line x1="80" y1="360" x2="680" y2="360" stroke="#30363d"/><text x="72" y="364" fill="#8b949e" font-size="10" text-anchor="end">0</text><text x="80" y="380" fill="#8b949e" font-size="10" text-anchor="middle">14</text><text x="230" y="380" fill="#8b949e" font-size="10" text-anchor="middle">18</text><text x="380" y="380" fill="#8b949e" font-size="10" text-anchor="middle">23</text><text x="530" y="380" fill="#8b949e" font-size="10" text-anchor="middle">28</text><text x="680" y="380" fill="#8b949e" font-size="10" text-anchor="middle">32</text><text x="380" y="400" fill="#8b949e" font-size="11" text-anchor="middle">peak memory (GB) →</text><text x="34" y="235" fill="#8b949e" font-size="11" text-anchor="middle" transform="rotate(-90 34 235)">decode tok/s →</text><circle cx="610.0" cy="291.7" r="24.7" fill="#818cf8" opacity="0.22"/><circle cx="610.0" cy="291.7" r="5" fill="#818cf8"/><text x="610.0" y="257.0" fill="#e6edf3" font-size="13" font-weight="700" text-anchor="middle">MLX-8bit</text><text x="610.0" y="272.0" fill="#8b949e" font-size="10.5" text-anchor="middle">8.63 bpw · max fidelity</text><circle cx="183.3" cy="233.3" r="18.3" fill="#fbbf24" opacity="0.22"/><circle cx="183.3" cy="233.3" r="5" fill="#fbbf24"/><text x="183.3" y="271.6" fill="#e6edf3" font-size="13" font-weight="700" text-anchor="middle">MLX-FP4</text><text x="183.3" y="286.6" fill="#8b949e" font-size="10.5" text-anchor="middle">4.88 bpw · compact</text><circle cx="236.7" cy="139.2" r="18.3" fill="#34d399" opacity="0.22"/><circle cx="236.7" cy="139.2" r="5" fill="#34d399"/><text x="236.7" y="110.9" fill="#e6edf3" font-size="13" font-weight="700" text-anchor="middle">FP4 + MTP bs3</text><text x="236.7" y="125.9" fill="#8b949e" font-size="10.5" text-anchor="middle">4.88 bpw · fastest</text></svg>