Feature Ideas

These graphics would be generated without a background, so we wouldn't need to remove the background afterwards with a background remover. When generating graphics, it's important that they have a transparent background, because graphics are meant to be overlayed on top of a canvas and combined with other graphics in order to create a design. Since their purpose is to be used as part of a bigger design, it's helpful if graphic elements are generated without a background. Currently, one AI tool that can generate PNG graphics with a transparent background is Creative Fabrica Spark. These graphics are automatically generated without a background, and there is no need to use a background remover in order to remove the background. Once these graphics are generated, we will then need to enlarge them by increasing their dimensions and making each one at least 300 DPI, which would make them suitable for printing. When enlarging graphics with a transparent background, it's important that the borders of the graphic remain crisp and there are no artifacts like fringing.

Images using flat colour or graphics suffer from compression artifacts. If you look at the edges of lines or shapes, you can see the compression. Images are downloaded as PNG - but have the visible compression of low quality JPEGs, and appear lower resolution as a result. Compared to competing AI art tools, it seems that they don't exibit the same artifacts - so something is happening along the image pipeline in the storage or the download of the generated images in SuperMachine. The issue becomes problematic for upscaling or professional use.

To make an animated picture

Like https://makeavideo.studio/

I really wish for the outpainting feature that's been created for both Stable DIffusion and Dall-E

Currently, the "Advanced Styling" options display as a list of names / buttons. Eg. "Expressionist", "Cubist", "by Andy Warhol" etc. It would be extremely helpful to have visual references or thumbnails to give a visual guide to all these styles. Without a art history major, or being super clued up on the art communities and digital artists online, it's very difficult to know what these styles all mean. Showing visual references would be a quick win in making this platform more user friendly - and into a better visual tool.

Quick way to make generations with previous prompts

Putting Pabbly / Integrately / Zapier etc all under this one area.

"Is it possible to add an extra user to the account? …For virtual assistants and such."

Inpainting is so usefull, wow i would love to see this implemented !!

Need to show the art of building a great prompt.

I feel like many of my prompts suffer from an excess of arms, and having an easy way to crop, blur or smooth over these would be outstanding! 💘

Hello please update API , there is no enought option. At least give the possibility to choose resolution 2K , 4K etc and the possibility to have consistent character

It's a minor detail, but worth considering ☺️

It would be nice to know what inputs are being used and be able to modify them - specifically the CFG Scale and the number of iterations. This could allow for more desirable outcomes. Thanks!