(Workflow Update)Qwen Image X Wan 2.2 Qwen3VL txt2img(20251030) (Patreon)

Based on all previous contents about Qwen Image. https://www.patreon.com/collection/1754194?view=expanded

I updated this workflow for image generation.

It mixed with Wan 2.2 Low Noise model for Img2Img, doing another additional sampling and upscale.

So overall, this workflow start from reference image to Qwen 3VL for Image to Text Prompt.

You can also enable ControlNet for same pose, but less creative.

Then it go to 2 sampling pass with Qwen Image, 2 Sampling with Wan 2.2.

All model links mentioned in tutorial video posts, too much in one post , I won't go through all in detail.

(Even I go through in detail, someone are skipping and not going through the content, repeat to ask anyway. XD )

If you need explainer check out the previous tutorials.

https://www.patreon.com/collection/1754194?view=expanded

Here are some examples that works with Character Lora.

Have Fun :)