Qwen-Image-2512-3bit-MLX

MLX-optimized 3-bit quantized version of Qwen-Image-2512 for Apple Silicon.

Quick Start

pip install mflux

mflux-generate-qwen \
  --model mlx-community/Qwen-Image-2512-3bit \
  --prompt "A photorealistic cat wearing a tiny top hat" \
  --steps 20

Model Details

• Base Model: Qwen/Qwen-Image-2512 (Dec 31, 2025)
• Quantization: 3-bit
• Framework: MLX (Apple Silicon optimized)
• Converted with: mflux 0.14.0

All Quantizations

• 8-bit (34GB)
• 6-bit (29GB)
• 5-bit (27GB)
• 4-bit (24GB)
• 3-bit (22GB)

License

Apache 2.0 (same as base model)