Stable Diffusion WebUI
Web interface for generating images from text
162.4k|stable|v1.10.1
Text-to-image generation — Generate images from text prompts with attention control and negative prompts.
Image editing modes — Inpainting, outpainting, and image-to-image transformation with loopback processing.
Model management — Load checkpoints on the fly, merge models, and switch VAE encoders from settings.
Neural upscaling — GFPGAN, RealESRGAN, ESRGAN, and SwinIR face restoration and image enhancement.
Parameter persistence — Automatically save generation settings in PNG metadata for reproducible results.
A full-featured Stable Diffusion interface that runs locally with extensive customization options. Supports txt2img, img2img, inpainting, and multiple upscaling models without requiring cloud services. Includes training tabs for embeddings and hypernetworks, checkpoint merging, and extensibility through community scripts.
Python 3.10.6, Git, NVIDIA/AMD GPU recommended (4GB VRAM minimum)
