stable-diffusion.cpp
C/C++ inference engine for diffusion image and video models
What it is
A ggml-based runtime, in the style of llama.cpp, that runs diffusion models such as Stable Diffusion, SDXL, FLUX, Qwen Image, Z-Image, Wan and LTX video without Python dependencies. It does text-to-image, image editing, and video generation, with LoRA, ControlNet, upscaling, and GGUF quantization, on CPU, CUDA, Vulkan, Metal, OpenCL and SYCL. Developers embed it as a library or use the sd-cli command and the bundled web UI to generate images on their own hardware.