# SD.Next

> An open-source, all-in-one WebUI for AI image and video generation built on Stable Diffusion, supporting dozens of advanced models across multiple platforms and hardware.

SD.Next is an open-source WebUI application for AI image and video generation, maintained by developer vladmandic on GitHub under the Apache License 2.0. It builds on the original Automatic1111 WebUI codebase and extends it with broad model support, cross-platform hardware compatibility, and unique performance features. The project is actively maintained, with the repository last pushed to in June 2026.

## What It Is

SD.Next is a self-hosted, all-in-one web interface for running Stable Diffusion and other generative AI models locally. It targets artists, researchers, and AI enthusiasts who want a powerful, flexible image and video generation environment without relying on cloud services. Users install it on their own hardware and access it through a browser-based UI that works on both desktop and mobile.

## Supported Models and Hardware

SD.Next supports a broad range of diffusion models documented in its model support pages. On the hardware side, it covers an unusually wide spectrum:

- **NVIDIA GPUs** via CUDA on Windows and Linux
- **AMD GPUs** via ROCm (Linux/Windows) and ZLUDA (Windows)
- **Intel Arc GPUs** via OneAPI/IPEX XPU on Windows and Linux
- **Any CPU/GPU** compatible with OpenVINO on Windows and Linux
- **DirectX-compatible GPUs** on Windows via DirectML
- **Apple M1/M2** on macOS via PyTorch MPS
- **ONNX/Olive** runtime support
- **Docker** container recipes for CUDA, ROCm, Intel IPEX, and OpenVINO

## Unique Technical Features

SD.Next ships several capabilities the README describes as not found in other WebUIs:

- **SDNQ**: A state-of-the-art quantization engine that supports pre-quantized models or on-the-fly quantization, claiming up to 4x VRAM reduction with minimal quality or performance impact
- **Balanced Offload**: Dynamically balances CPU and GPU memory to run larger models on limited hardware
- **Captioning**: Supports 150+ OpenCLIP models, WaifuDiffusion and DeepDanbooru taggers, and 25+ built-in VLMs
- **Image Processing**: A full color-grading and image correction suite built into the interface
- Localization into approximately 15 languages with support for multiple UI themes

## Setup Path

SD.Next uses a built-in installer with automatic updates and dependency management. The quick-start path is a `git clone` followed by running `webui.sh` (Linux/Mac), `webui.bat` (Windows), or `webui.ps1` (PowerShell). Platform-specific guides cover WSL, Intel Arc, DirectML, OpenVINO, ONNX/Olive, ZLUDA, AMD ROCm, macOS, NVIDIA, and Docker deployments. An install walkthrough video is linked from the README.

## Project Status and Community

The repository was created in December 2022 and shows active development, with the last commit pushed in June 2026. The project has accumulated over 7,100 GitHub stars and 562 forks. Community support runs through a Discord server and GitHub Issues and Discussions. Documentation is hosted separately at the project's GitHub Pages docs site, and a DeepWiki badge indicates third-party documentation indexing.

## Features
- AI image generation
- AI video generation
- Stable Diffusion support
- Multi-model support
- SDNQ quantization engine
- Balanced CPU/GPU offload
- Image captioning (150+ OpenCLIP models)
- WaifuDiffusion and DeepDanbooru tagging
- 25+ built-in VLMs
- Image processing and color-grading suite
- Desktop and mobile UI
- ~15 language localizations
- Multiple UI themes
- Built-in installer with auto-updates
- Docker support
- NVIDIA CUDA support
- AMD ROCm support
- Intel Arc/OneAPI support
- OpenVINO support
- DirectML support
- Apple M1/M2 MPS support
- ONNX/Olive runtime support

## Integrations
Stable Diffusion, Diffusers, PyTorch, CUDA, ROCm, ZLUDA, OneAPI/IPEX XPU, OpenVINO, DirectML, Apple MPS, ONNX/Olive, Docker, OpenCLIP, WaifuDiffusion, DeepDanbooru

## Platforms
WINDOWS, MACOS, LINUX, WEB, API, CLI

## Pricing
Open Source

## Links
- Website: https://vladmandic.github.io/sdnext-docs/
- Documentation: https://vladmandic.github.io/sdnext-docs/
- Repository: https://github.com/vladmandic/sdnext
- EveryDev.ai: https://www.everydev.ai/tools/sdnext