Learning how to change the aspect ratio of an image with AI is very easy, but unlike cameras and design software, there’s a few more considerations.

AI models “learn” using square 1:1 aspect ratio images, so when we ask for a picture back that isn’t a square, this can result in glitches like multiple heads on a person, the same person twice in an image, or a crash: causing images to render poorly or not at all. Yikes.

two men and a burning building

But hey, sometimes double images are kinda cool

To get the best images, we should multiply the size we want by 64 pixels. We’re definitely not asking our users to crunch this math every time. No way.

Instead, we created shortcuts that render the most compatible dimensions without burdening you with the details. Use these instead:

Use these shortcut commands for best compatibility with all AI models. Behind the scenes, we are using exact multiples of 64 pixels. 


/render /wide a man in front of a burning building

/render /tall a man in front of a burning building

Deprecated: /size

You can still manually use /render /size:512×512, but this may result in unusual behavior. If you must use size, multiply by 64 for best results. Example: 704 is better than 700, because (64x11=704)

And now for a bunch of cool wide renders from our test session, because we wouldn’t want these imaginary burned buildings to go to waste.

Want to make images like these?  Check out Analog style


You may also like

More in Guide