Can I use this workflow to edit multiple images at once?

Yes, you can process multiple images by setting up batch inputs in the LoadImage node and running the workflow for each image.

How do I customize the output image size?

To customize the image size, delete the Get Image Size node and manually set the dimensions in the ImageScaleToTotalPixels node.

What should I do if the image output is not as expected?

Ensure that your prompts are clear and specific. You can also adjust the sampling settings in the SamplerCustomAdvanced node for better results.

Is it possible to add a second image input?

Yes, you can enable the second image input by using Ctrl + B to switch all nodes in Bypass mode to normal mode.

OmniGen2 Image Edit - ComfyUI Workflow

Back

The OmniGen2 Image Edit workflow is a powerful tool designed to facilitate image editing using natural language instructions. At its core, this workflow leverages the OmniGen model to transform images based on user-defined prompts. It integrates a variety of nodes such as CLIPTextEncode for text processing, VAEDecode for image decoding, and SaveImage for outputting the final result. This setup allows users to perform complex image transformations with simple text commands, making it highly accessible for both novice and advanced users.

Technically, the workflow begins by loading the necessary models using nodes like CLIPLoader and UNETLoader. It then processes the input image through nodes such as VAEEncode and ReferenceLatent to prepare it for transformation. The use of RandomNoise and BasicScheduler nodes adds controlled variations, while the DualCFGGuider and SamplerCustomAdvanced nodes ensure high-quality sampling and image generation. This workflow is particularly useful for artists and designers looking to quickly iterate on visual concepts or for anyone needing to make precise image edits without extensive manual work.