# QR Code Monster LoRa for Qwen Image 2.1 Edit — retained record

> Retained reddit record from reddit:StableDiffusion.

Canonical: https://highsignal.app/data/records/4faedf54a37be400

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:StableDiffusion)  published 2026-10-04 16:18 ingested 2026-10-05 02:39

[original source ↗](https://www.reddit.com/r/StableDiffusion/comments/1wxjmi0/qr_code_monster_lora_for_qwen_image_21_edit/)

## retained text



```
For those who are unaware, there was an old Controlnet for old SD models that was quite popular here back in those times, that would make cool almost Optical Illusion-like images. This is the same concept but instead I'm using I2I Image edit model where you provide an input image and ask the model to generate the scene of your choice.

I've been very pleased with the results from the Qwen Image 2.1 version. It was trained on 25 high quality image pairs and it works quite well, even when using human subjects and depicting scenes not included the dataset. It can do QR Codes but it was not the main focus of the dataset so it can be hit or miss as far as the actual usability/scanability of the codes.

The suggested prompt is exactly as captioned in the dataset: `Transform the entire image into` <your prompt here> `while preserving the outlines and shapes of the original image`

It seems to work well with complex prompts and simple ones as well.

You can download it on Civit [here](https://civitai.red/models/2432921). All images/videos should include a workflow which is basically just a pixel drift fix workflow from Ausboss with a few added touches and using a viggle Turbo LoRa. I'm mainly posting this here as I would love to see what you create!
```



document fetched 2026-10-05 00:29 · canonical https://reddit.com/r/StableDiffusion/comments/1wxjmi0/qr_code_monster_lora_for_qwen_image_21_edit
