Homebrew Tap for JANG

JANG — Adaptive Mixed-Precision Quantization for Apple Silicon. The GGUF equivalent for MLX.

Install

brew tap jjang-ai/jangq
brew install jang

Usage

# Convert any HuggingFace model
jang convert Qwen/Qwen3.5-35B-A3B -p 4

# 2-bit for extreme compression
jang convert model -p 2

Links

• GitHub
• PyPI
• HuggingFace Models
• MLX Studio