Qwen-Text-to-Image-Open-Door-qwen image 2512

Fan benefits: Invitation code: rh v1058 Click on the avatar in the upper right corner, enter the invitation code in the "Invitation Code" section to get 1000 RH coins. Logging in daily also earns you 100 coins.
Usage instructions
1. Switch: Turning it on uses the prompt words derived from the uploaded image, turning it off uses your own prompt words.
2. Upload reference image [mainly used when you don't have prompt words, using the image to derive prompt words to regenerate the image]. When the switch is off, this function is not effective.
3. Aspect ratio selection: 1 for vertical image 3:4, 2 for horizontal image 4:3, 3 for vertical image 9:16, 4 for horizontal image 16:9.
4. Provide your own prompt words. When the switch is on, you can fill in prompt words for specific clothing of the character, and it will also work in reference image mode, e.g., wearing a white sweater.
      When the switch is off, you need to fill in complete image prompt words.
Note: The final output is a grayscale image, which requires local decoding or web decoding.
Web decoding: http://smalljj.top/
Local decoding ComfyUI node download link for mainland China: https://pan.quark.cn/s/0de658cdeeb6
Local decoding ComfyUI node download link for other countries: https://drive.google.com/drive/folders/1cmn9oK8Ojj7gZm5eZaPGP8bNYLaZvX7e?usp=drive_link
The entire image generation process takes about 3 minutes.
If you need to contact me or encounter any issues with the application, you can search the official account: ai小健健, to reach me for support.