Key Features

What you can do

Multi-Model Support

Supports multiple diffusion models with built-in downloaders for CivitAI and HuggingFace, including Stable Diffusion XL, Stable Diffusion 3.x, Stable Cascade, FLUX.1, HiDream, and LCM.

Multi-Platform Hardware Acceleration

Automatic detection and tuning for NVIDIA CUDA, AMD ROCm, Intel Arc/IPEX XPU, DirectML, OpenVINO, ONNX+Olive, and ZLUDA across Windows, Linux, and macOS.

Built-in Processing Tools

Includes tools for text/image/batch/video processing, ControlNet, Detailer, styles, wildcards, outpainting, and reprocessing workflows.

CLI and API Support

Provides command-line interface tools and APIs for image preprocessing, visual question answering, text-to-image generation, benchmarking, and model conversion.

Model Optimization

Supports model quantization methods (SDNQ, BitsAndBytes, Optimum-Quanto, TorchAO) and compile backends (Triton, StableFast, DeepCache, OneDiff, TeaCache) for performance tuning.

Localization and UI Modes

Offers full localization in 10 languages and multiple user interface modes including Standard and Modern.