かみもと · 本地AI视频工作流·· 2026-10-18AI score42
MOSS-Image Hands-On in ComfyUI: A 5B Model Running Anime and Realistic Styles
【MOSS-Image】新作画像生成AI「MOSS-Image」をComfyUIで検証してみた(実写・アニメ絵)
AI briefing
MOSS-Image, a 5B-parameter DiT image generation model released by the OpenMOSS team, is now usable in ComfyUI itself. The author used an RTX 4070 (12GB VRAM) with INT8 weights to test both anime and realistic styles.
Open original note.com
Image GenerationComfyUI & WorkflowsTutorials & Practice#Tutorial#Image generation#ComfyUI & workflows