In the gallery, when viewing an image detail on phone, the image is stretched out.

To reproduce:

1. generate a square image with large enough dimensions (512x512 is enough on my phone, but you may try 1024x1024)
2. go to the gallery and find the image, tap it to see the detail
3. the image is distorted

Only selecting source image for now, with future support for the following in mind:

• sketching
• inpainting

by model, by lora(s), by words in prompt, etc

vanilla Stable diffusion 1.5 is way too old and someone would get the wrong idea of the capabilities of the website if it's the first they use. Suggested you switch to using Stable Cascade as the default, or alternatively a generalist SDXL model you might prefer, like Albedo

https://github.com/amiantos/AI-Horde-Styles-Previews

Would be nice if this repo is added to the website and also would appreciate a link (and potentially also an explanation) of the AI Horde as well as that is what can convert some users into workers.

Check out artbot for inspiration

Loading the full image just for the thumbnail is way too slow.

After generating an image on phone, there's the download icon which shows an overlay with Full image, Download etc.

• Full image should have a non breaking space
• The buttons are clickable even if the overlay is not visible
• Perhaps use a different icon than download?
• The buttons cut into the text below them

Example: All Disney Princess XL LoRA Model from Ralph Breaks the Internet

Do an analysis of the order and styling of stuff on the generate page so that the most important stuff is visible