Instructions to use xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp 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("xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp")
config = load_config("xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp")

# 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 xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp"

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": "xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp 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 "xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp"

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 xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp

Run Hermes

hermes

OpenClaw new

How to use xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp"

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 "xunkutech-ai/Qwythos-9B-v2-MLX-bf16-mtp" \
  --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"

4-bit quantified version

by gao204 - opened 20 days ago

Discussion

gao204

20 days ago

Can you provide a 4-bit quantified version? with vision support please~， Thank your efforts.

xunkutech-ai

Owner 20 days ago

Hugging Face CLI

You can convert this model to any quantization precision you want using oMLX's built-in oQ quantization tool. Just make sure to keep the vision model and MTP tensors when converting (this is mentioned in the model card).

I also recommend using quantization levels oQ6 or higher — lower precision (e.g., oQ4) tends to introduce noticeable errors in instruction-following and tool-use capabilities.

The oQ6 and oQ8 quantized versions are now available at:

You can download and use them directly.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment