moonbox-ai/Moebius-MLX

Moebius MLX

MLX builds of Moebius, made by Moonbox for Apple Silicon Macs. Each precision is a folder; keep a folder's files together. Only the parts listed below are converted: the rest of the pipeline comes from the original repository unless the section below says otherwise.

Files

The last column is how far each build is from the original PyTorch model on the same inputs (see the section below for what was measured).

Precision	File	Size	sha256	Checked against the original
bf16	bf16/temb.npz	5.1 MB	b5b4a28cb0ba	PSNR 32.2 dB mean / 27.2 worst, LPIPS <= 0.050
bf16	bf16/unet.safetensors	476.8 MB	13440a538271	PSNR 32.2 dB mean / 27.2 worst, LPIPS <= 0.050
bf16	bf16/vae.safetensors	167.3 MB	20493a16c12c	PSNR 32.2 dB mean / 27.2 worst, LPIPS <= 0.050
fp16	fp16/temb.npz	5.1 MB	b5b4a28cb0ba	PSNR 41.0 dB mean / 33.9 worst, LPIPS <= 0.008
fp16	fp16/unet.safetensors	476.8 MB	e75b9e1549e4	PSNR 41.0 dB mean / 33.9 worst, LPIPS <= 0.008
fp16	fp16/vae.safetensors	167.3 MB	81593a7667f4	PSNR 41.0 dB mean / 33.9 worst, LPIPS <= 0.008
8bit	8bit/temb.npz	5.1 MB	b5b4a28cb0ba	PSNR 37.4 dB mean / 33.0 worst, LPIPS <= 0.022
8bit	8bit/unet.safetensors	322.2 MB	57339be34c7f	PSNR 37.4 dB mean / 33.0 worst, LPIPS <= 0.022
8bit	8bit/vae.safetensors	167.3 MB	c7eaffa5125e	PSNR 37.4 dB mean / 33.0 worst, LPIPS <= 0.022
6bit	6bit/temb.npz	5.1 MB	b5b4a28cb0ba	PSNR 32.9 dB mean / 26.4 worst, LPIPS <= 0.066
6bit	6bit/unet.safetensors	281.0 MB	f4d5da54ee83	PSNR 32.9 dB mean / 26.4 worst, LPIPS <= 0.066
6bit	6bit/vae.safetensors	167.3 MB	4d8ad4179f1c	PSNR 32.9 dB mean / 26.4 worst, LPIPS <= 0.066
4bit	4bit/temb.npz	5.1 MB	b5b4a28cb0ba	below threshold (experimental): PSNR 22.4 dB mean / 16.7 worst, LPIPS <= 0.380
4bit	4bit/unet.safetensors	239.8 MB	775f131ac8b5	below threshold (experimental): PSNR 22.4 dB mean / 16.7 worst, LPIPS <= 0.380
4bit	4bit/vae.safetensors	167.3 MB	2c9ec84a4993	below threshold (experimental): PSNR 22.4 dB mean / 16.7 worst, LPIPS <= 0.380

How to run

Moebius (the pretrained general model) with the PixelHacker VAE, converted for the MLX port in Moonbox's Moebius engine. Each folder holds:

File	What
unet.safetensors	the 0.22 B UNet: BatchNorms, norm affines and the constant conditioning folded into the weights; 1x1 convs and attention projections stored as [in, out] matrices (MLX affine-quantized, group 64, in the 8bit / 6bit / 4bit folders: <name>.q, .scales, .biases)
vae.safetensors	PixelHacker's VAE (fp16 in the quantized folders)
temb.npz	silu(time_embedding(t)) for t = 0..999

The layout is the one moebius_od/mlx_unet.py and mlx_vae.py in Moonbox read (NHWC, 512x512, 20 DDIM steps, CFG 2, as upstream's README). That runtime isn't part of this repository.

How they were checked. Through the same MLX engine against the float32 weights, removing the main subject of 8 photos (seed 0): PSNR and LPIPS of the 512 px output. Pass: LPIPS <= 0.15 on every image. The MLX port itself was checked against the official PyTorch pipeline when the engine was written.

Precision	Size	PSNR vs float32 (mean / worst)	LPIPS (worst)	Status
fp16/	649 MB	41.0 / 33.9 dB	0.008	passes
bf16/	649 MB	32.2 / 27.2 dB	0.050	passes
8bit/	495 MB	37.4 / 33.0 dB	0.022	passes
6bit/	453 MB	32.9 / 26.4 dB	0.066	passes
4bit/	412 MB	22.4 / 16.7 dB	0.380	below threshold, experimental

About 4-5 s per inpainting (20 steps, 512 px) on an Apple M5.

How these were made

From Moebius at revision cd01f47fb6: the original weights loaded in MLX and stored as bf16 or fp16, or quantized with MLX's affine quantization (8bit, 6bit, 4bit, group size 64; layers MLX can't quantize, such as convolutions and norms, stay bf16). Every build was checked against the original before upload; the sha256 column above is what the files must hash to.

License

Apache-2.0. These files are a derivative of the original model and carry its license: it is copied in this repository, and it governs your use of these files. The original model card follows unchanged.

---

Original model card

《Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance》
Project Page: https://hustvl.github.io/Moebius
Paper: https://arxiv.org/abs/2606.19195
Code: https://github.com/hustvl/Moebius
weight: https://huggingface.co/hustvl/Moebius/tree/main