🎨 MOSS-Image Character and Style Catalog Fully Online

The 'Neuronaut' channel compiled a digest of ~20 free resources for the OpenMOSS open-source model MOSS-Image. The MOSS-Image-Character-Catalog on Hugging Face is now searchable: 33,935 characters and 314 styles with previews and full prompts (v0.1.2 from November 4, Shuffle mode). A layout editor, SwarmUI-LayoutEditor, has been released for scenes, featuring JSON prompt export.

🌍 Within three weeks, a third-party stack comparable to Stable Diffusion has emerged around MOSS-Image: style catalogs, layout editors, samplers for ComfyUI, INT4 compression for WebGPU, and LoRA on Civitai. The model with strong text rendering has become a public platform.

👤 Free and GPU-free: style prompts from the catalog, layouts in the editor, and the 4-step INT4 version (7.5 GB) can be run directly in the browser.

Source 1: https://huggingface.co/spaces/hide1/MOSS-Image-Character-Catalog Source 2: https://github.com/jambavant/SwarmUI-LayoutEditor