Omnigen2 example.

Author: comfyanonymous

README.md

@@ -60,6 +60,8 @@ Here are some more advanced examples:
 
 [Edit/InstructPix2Pix Models](edit_models)
 
+[Omnigen2](omnigen)
+
 [Stable Video Diffusion](video)
 
 [Mochi](mochi)

omnigen/README.md

@@ -0,0 +1,24 @@
+# Omnigen 2
+
+[Omnigen 2](https://github.com/VectorSpaceLab/OmniGen2) is a model that can be used to edit images with text prompts.
+
+## Files to Download
+
+You will first need:
+
+[omnigen2_fp16.safetensors](https://huggingface.co/Comfy-Org/Omnigen2_ComfyUI_repackaged/blob/main/split_files/diffusion_models/omnigen2_fp16.safetensors) goes in: ComfyUI/models/diffusion_models/
+
+[qwen_2.5_vl_fp16.safetensors](https://huggingface.co/Comfy-Org/Omnigen2_ComfyUI_repackaged/blob/main/split_files/text_encoders/qwen_2.5_vl_fp16.safetensors) goes in: ComfyUI/models/text_encoders/
+
+[ae.safetensors](https://huggingface.co/Comfy-Org/Omnigen2_ComfyUI_repackaged/blob/main/split_files/vae/ae.safetensors), this is the flux VAE that you might already have, it goes in: ComfyUI/models/vae/
+
+## Workflows
+
+This is a basic workflow using an image as a character reference. For multiple image inputs chain ReferenceLatent nodes together
+
+![Example](omnigen2_example.png)
+
+You can load this image in [ComfyUI](https://github.com/comfyanonymous/ComfyUI) to get the full workflow.
+
+You can find the input image [here](../chroma/fennec_girl_sing.png)
+