Instructions to use GreenBitAI/Qwen3.8-Flash-Next-4bit-paged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GreenBitAI/Qwen3.8-Flash-Next-4bit-paged 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("GreenBitAI/Qwen3.8-Flash-Next-4bit-paged")
config = load_config("GreenBitAI/Qwen3.8-Flash-Next-4bit-paged")

# 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 GreenBitAI/Qwen3.8-Flash-Next-4bit-paged with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "GreenBitAI/Qwen3.8-Flash-Next-4bit-paged"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "GreenBitAI/Qwen3.8-Flash-Next-4bit-paged"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent

How to use GreenBitAI/Qwen3.8-Flash-Next-4bit-paged 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 "GreenBitAI/Qwen3.8-Flash-Next-4bit-paged"

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 GreenBitAI/Qwen3.8-Flash-Next-4bit-paged

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use GreenBitAI/Qwen3.8-Flash-Next-4bit-paged with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "GreenBitAI/Qwen3.8-Flash-Next-4bit-paged"

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 "GreenBitAI/Qwen3.8-Flash-Next-4bit-paged" \
  --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"

Qwen3.8-Flash-Next-4bit-paged

Expert-paged build of Vontra/Qwen3.8-Flash-Next-MLX-4bit. The weights that are read a fraction at a time live in their own containers, so a machine loads what it needs rather than all of it.

file	size	holds
`model.safetensors`	3.80 GiB	resident weights
`experts.bin`	70.31 GiB	routed experts
`ple-q4.rows`	29.80 GiB	n-gram table

Total 103.94 GiB — the same as the source. The bytes moved, they were not copied.

from gbx_lm.utils import load
model, tokenizer = load("GreenBitAI/Qwen3.8-Flash-Next-4bit-paged")

Where the weights fit they are filled from experts.bin and the model runs the stock path at stock speed; where they do not, they stream from disk. Reading the machine decides that, not a flag.

Checked at build time, while the source checkpoint was still there to compare against:

PASS bit-identical logits — exact on 5 prompt(s) to 160 tokens; 1 longer differ by at most 6.8, same greedy token throughout
PASS layer-wise vs resident — 48 layers x 2 draws exact, 88.89 GiB peak for this gate

Quantization, tokenizer, chat template and licence are unchanged from Vontra/Qwen3.8-Flash-Next-MLX-4bit.

Downloads last month: 64

Safetensors

Model size

1B params

Tensor type

BF16

U32

I64

MLX

Hardware compatibility

4-bit

Model tree for GreenBitAI/Qwen3.8-Flash-Next-4bit-paged

Base model

Qwen/Qwen3.8-Flash-Next

Quantized

(135)

this model