moss-image

MOSS-Image for Stable Diffusion WebUI Forge Neo: text-to-image with sharp English and Chinese text, img2img and layouts by boxes, with the SFT and Base models

Imagination Studio by Keys (x.com/u1tra_instinct) + Sam Wasserman / Wasserman Productions (wassermanproductions.com · x.com/SamJWasserman), powered by TensorFold: a local all-in-one AI film studio for NVIDIA DGX Spark and Apple silicon, with Wasserman's Filmmaker Suite built in.

Run MOSS-Image locally on Apple silicon Macs. Staged Metal/MPS inference, prompt caching, FP16 and Gradio UI. Mac 本地文生图。

Web UI for training LoRAs on Anima, Krea 2, MOSS-Image, SDXL, FLUX, SD3.5 and SD1.5. Tuned for speed: per-block torch.compile, flash-attn, fp16 accumulation and a batched ProdigyPlus step (1.5-2x faster adds up!). Built on kohya-ss sd-scripts and musubi-tuner, with its own MOSS-Image trainer.

MOSS-Image (OpenMOSS's bilingual text-rendering DiT) on Apple Silicon with MLX: exact port, 9 GB peak

🌿 Pocket MOSS Image — run the full MOSS-Image 9B encoder in pocket-sized VRAM: packed INT6 weights + streamed fused INT8 DiT, ~6.6 GB GPU peak, no retraining.

ComfyUI custom nodes: visual layout editor for MOSS-Image — draw boxes on a canvas and compile the official summary/items layout JSON.

DeepSeek Harness plugin for MOSS-Image: switch DiT (BF16/FP8) and Text Encoder (BF16/FP8) precision combos without downloading or replacing weights.

DyPE-M dynamic position extrapolation for MOSS-Image - render past the 4MP training range without retraining

MOSS-Image Base/SFT for ComfyUI: official pipeline, separate DiT/TE/VAE loaders, ready-to-run workflows.

A detailed diagram laying out the full MOSS-Image architecture as released by OpenMOSS at https://github.com/OpenMOSS/MOSS-Image.

神都猫 MOSS 单图编辑节点 | AI-Toolkit-aligned MOSS single-image editing for ComfyUI, bilingual guides and v2 LoRA workflows

A prompt enhancing plugin for newly released t2p model MOSS Image

ComfyUI 提示词扩写节点：通过兼容 Chat Completions 的 API，将自然语言需求编译成 MOSS 布局 JSON 或自然语言提示词。

Local MOSS-Image text-to-image studio for Windows and NVIDIA RTX GPUs, with CUDA and WeeLLM streaming for low-VRAM systems.

Inference Krea-2, Qwen Image 2.x , MOSS-Image, Flux 3 Image Klein with ComfyUI

Run MOSS-Image-SFT (5B, INT4) locally on Apple Silicon Macs, with an MCP server: text-to-image with reliable text, bounding-box layouts, inpainting, cutouts. ~30 s/image on a base M4.

Run MOSS-Image text to image in Forge Neo. Model, text encoder, VAE and one preset. Tested on Forge Neo neo-2.30.1. By Stable Yogi.

ComfyUI node that feeds a reference image through the Qwen3.5-9B vision tower and injects its embeddings into the text prompt (MOSS-Image conditioning)

MOSS-Image SFT on Runpod: curated ComfyUI workflows and a queue API, with pinned models and reproducible deployment.