Image-Text-to-Text
MLX
Safetensors
English
qwen3_5
mlx-vlm
apple-silicon
metal
mixed-precision
quantized
qwen
qwen3
qwen3.6
multimodal
vision
mtp
speculative-decoding
gated-deltanet
mamba
ssm
linear-attention
uncensored
abliterated
refusal-removed
aeon
aeon-7
m4-pro
on-device
conversational
mxfp4
fp4
4-bit precision
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
- MLX
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
- Pi
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"
File size: 4,883 Bytes
da35640 | 1 | <svg viewBox="0 0 820 440" xmlns="http://www.w3.org/2000/svg" font-family="system-ui,-apple-system,Segoe UI,Roboto,sans-serif"><rect width="820" height="440" rx="14" fill="#0d1117"/><rect x="1" y="1" width="818" height="438" rx="13" fill="none" stroke="#30363d"/><text x="28" y="40" fill="#e6edf3" font-size="21" font-weight="700">Per-category decode throughput — baseline vs MTP</text><text x="28" y="63" fill="#8b949e" font-size="12.5">FP4 · MTP block-size 3 (lossless) · decode tok/s, single stream · M4 Pro 48 GB</text><rect x="540" y="31" width="11" height="11" rx="2" fill="#fbbf24"/><text x="556" y="40" fill="#8b949e" font-size="11.5">baseline</text><rect x="618" y="31" width="11" height="11" rx="2" fill="#34d399"/><text x="634" y="40" fill="#8b949e" font-size="11.5">+ MTP bs3</text><line x1="70" y1="360" x2="790" y2="360" stroke="#30363d"/><text x="62" y="364" fill="#8b949e" font-size="10" text-anchor="end">0</text><line x1="70" y1="277" x2="790" y2="277" stroke="#30363d"/><text x="62" y="281" fill="#8b949e" font-size="10" text-anchor="end">10</text><line x1="70" y1="193" x2="790" y2="193" stroke="#30363d"/><text x="62" y="197" fill="#8b949e" font-size="10" text-anchor="end">20</text><line x1="70" y1="110" x2="790" y2="110" stroke="#30363d"/><text x="62" y="114" fill="#8b949e" font-size="10" text-anchor="end">30</text><rect x="91.5" y="234.0" width="36.0" height="126.0" rx="3" fill="#fbbf24"/><text x="109.5" y="228.0" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">15.1</text><rect x="132.5" y="149.7" width="36.0" height="210.3" rx="3" fill="#34d399"/><text x="150.5" y="143.7" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">25.2</text><text x="130.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Code</text><text x="130.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.67x · 85% acc</text><rect x="211.5" y="236.2" width="36.0" height="123.8" rx="3" fill="#fbbf24"/><text x="229.5" y="230.2" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">14.8</text><rect x="252.5" y="146.1" width="36.0" height="213.9" rx="3" fill="#34d399"/><text x="270.5" y="140.1" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">25.7</text><text x="250.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Math</text><text x="250.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.73x · 90% acc</text><rect x="331.5" y="237.0" width="36.0" height="123.0" rx="3" fill="#fbbf24"/><text x="349.5" y="231.0" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">14.8</text><rect x="372.5" y="163.9" width="36.0" height="196.1" rx="3" fill="#34d399"/><text x="390.5" y="157.9" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">23.5</text><text x="370.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Reasoning</text><text x="370.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.59x · 82% acc</text><rect x="451.5" y="237.2" width="36.0" height="122.8" rx="3" fill="#fbbf24"/><text x="469.5" y="231.2" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">14.7</text><rect x="492.5" y="166.8" width="36.0" height="193.2" rx="3" fill="#34d399"/><text x="510.5" y="160.8" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">23.2</text><text x="490.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Creative</text><text x="490.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.57x · 74% acc</text><rect x="571.5" y="235.0" width="36.0" height="125.0" rx="3" fill="#fbbf24"/><text x="589.5" y="229.0" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">15.0</text><rect x="612.5" y="151.3" width="36.0" height="208.7" rx="3" fill="#34d399"/><text x="630.5" y="145.3" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">25.0</text><text x="610.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Knowledge</text><text x="610.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.67x · 84% acc</text><rect x="691.5" y="232.8" width="36.0" height="127.2" rx="3" fill="#fbbf24"/><text x="709.5" y="226.8" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">15.3</text><rect x="732.5" y="178.2" width="36.0" height="181.8" rx="3" fill="#34d399"/><text x="750.5" y="172.2" fill="#e6edf3" font-size="11.5" font-weight="600" text-anchor="middle">21.8</text><text x="730.0" y="380" fill="#e6edf3" font-size="12.5" text-anchor="middle" font-weight="600">Chat</text><text x="730.0" y="397" fill="#34d399" font-size="10.5" text-anchor="middle">1.43x · 67% acc</text><text x="70" y="98" fill="#8b949e" font-size="10">tok/s</text></svg> |