ComfyUI Tutorial Series: Mastering Stable Diffusion AI - ComfyUI Tutorial Series: Ep21 - How to Use OmniGen in ComfyUI

Welcome to Episode 21 of our ComfyUI tutorial series! In this video, we’ll cover OmniGen, a new image-generation model that enables a range of creative possibilities using multi-modal prompts within ComfyUI. We’ll guide you through setting up OmniGen, installing necessary dependencies, and handling common installation issues. This model can generate complex images, combine multiple prompts, and apply different artistic styles—all in one workflow. Note that OmniGen requires a high amount of VRAM (around 16GB), so make sure your system is equipped. Unlock exclusive perks by joining our channel: https://www.youtube.com/channel/UCmMbwA-s3GZDKVzGZ-kPwaQ/join Get the workflows and instructions from discord https://discord.gg/gggpkVgBf3 ----- Go to Manger, Custom Nodes Manager and search for ComfyUI-OmniGen Install ComfyUI-OmniGen by the author 1038lab Restart ComfyUI ----- Read more about OmniGen https://github.com/VectorSpaceLab/OmniGen Check the OmniGen ComfyUI Node https://github.com/1038lab/ComfyUI-OmniGen If you want to manually download the model you can do it from this page, it has 15.5 GB model.safetensors https://huggingface.co/Shitao/OmniGen-v1/tree/main place it in the ..ComfyUI\models\LLM\OmniGen-v1 You can also try to see if it works faster the smaller model model-fp8_e4m3fn.safetensors that has only 3.88 GB and place it in same folder but rename it to model.safetensors so move the bigger model in different folder and test with small model.safetensors in the folder https://huggingface.co/silveroxides/OmniGen-V1/tree/main ---- #comfyui #omnigen #comfyuiseries Chapters 00:00 Introduction OmniGen 00:31 OmniGen Comfy UI Install 05:54 OmniGen Tests 11:04 Upscale OmniGen Images with Flux 12:12 Smaller OmniGen Model