...It injects control models dynamically without requiring them to be merged into the base Stable Diffusion checkpoint. Users can guide generation with structures such as edges, depth, poses, line art, segmentation maps, and other preprocessed inputs. Multiple ControlNet units can influence the same generation with independent weights and timing ranges. The extension supports txt2img, img2img, inpainting, masks, batch workflows, and High-Res Fix. It also works with T2I-Adapter and several related control architectures. Web and internal APIs allow other applications or extensions to control its generation features programmatically.