Instructions to use 0XARTEX/artex-coder-7b-mlx-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 0XARTEX/artex-coder-7b-mlx-4bit with MLX:

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

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("0XARTEX/artex-coder-7b-mlx-4bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use 0XARTEX/artex-coder-7b-mlx-4bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "0XARTEX/artex-coder-7b-mlx-4bit"

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": "0XARTEX/artex-coder-7b-mlx-4bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

MLX LM

How to use 0XARTEX/artex-coder-7b-mlx-4bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "0XARTEX/artex-coder-7b-mlx-4bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "0XARTEX/artex-coder-7b-mlx-4bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "0XARTEX/artex-coder-7b-mlx-4bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Hermes Agent

How to use 0XARTEX/artex-coder-7b-mlx-4bit 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 "0XARTEX/artex-coder-7b-mlx-4bit"

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 0XARTEX/artex-coder-7b-mlx-4bit

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use 0XARTEX/artex-coder-7b-mlx-4bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "0XARTEX/artex-coder-7b-mlx-4bit"

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 "0XARTEX/artex-coder-7b-mlx-4bit" \
  --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"

Artex Coder 7B · MLX 4-bit

4-bit MLX build of Artex Coder 7B for Apple Silicon. Artex is a concise, accurate, to-the-point coding assistant fine-tuned from Qwen2.5-Coder-7B-Instruct. It answers in the user's language.

Usage

pip install mlx-lm
mlx_lm.generate --model 0XARTEX/artex-coder-7b-mlx-4bit --prompt "Write a Python function that checks if a string is a palindrome."

from mlx_lm import load, generate

model, tokenizer = load("0XARTEX/artex-coder-7b-mlx-4bit")
messages = [{"role": "user", "content": "How do I reverse a list in JavaScript?"}]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
print(generate(model, tokenizer, prompt=prompt, max_tokens=512))

Build


Source	0XARTEX/artex-coder-7b (bf16)
Quantization	4-bit affine, group size 64 (4.5 bits per weight)
Scales / biases dtype	float16
Size	4.0 GB

Scales are stored as float16, not bfloat16. Apple M1 has no hardware support for bfloat16, and an earlier internal build with bfloat16 scales decoded about 19% slower on the same file size.

Speed (measured)

Apple M1 Max, 64 GB, mlx-lm, single stream, three coding prompts, 400 max tokens:

Build	Decode
bfloat16 scales (internal, unreleased)	~53 tok/s
float16 scales (this repo)	~63 tok/s

Peak memory for one loaded model is about 4.4 GB. Your numbers will vary with chip, prompt and context length.

Limitations

No standardized benchmark (HumanEval, MBPP) has been run yet.
Like any LLM, Artex can produce incorrect or insecure code. Review its output before using it.