Edit specific image areas (e.g., object removal, detail refinement) while preserving overall composition.
This ComfyUI workflow offers rapid image editing with multi-image input and a multi-LoRA loader, automatically scaling images for optimal quality. It supports Qwen and lighting LoRAs, includes a prompt stasher, and an image comparison slider, with specific sampler settings for best results and custom image sizing recommendations.
OmniGen2 is a powerful multimodal generative model with a dual-path architecture (3B text, 4B image diffusion) for efficient, specialized optimization. It offers visual understanding, high-fidelity text-to-image generation, instruction-guided image editing, and context-aware visual output, with excellent detail preservation and text generation in images.
HiDream E1.1 is a simple ComfyUI workflow that uses the HiDream-E1 model for super-resolution, enhancing low-resolution images into high-definition outputs, particularly effective for anime and illustrations.
This ComfyUI workflow removes clothing from a person in an image, using a segmentation model to identify clothing and body areas, flux and redraw nodes for removal, and optional prompts for refinement.
Filters