The plug-in is still in beta.
Former Adobe's Director of Design Prototyping Christian Cantrell has shared some new functions of the Stable Diffusion plug-in for Photoshop. While still in beta, it now allows inpainting a rectangular area with the text-to-image tool's AI without masking, replacing the selected image with the prompt you give the model. If you'd like to work with a more flexible area, you still have to use masks.
Another change in the new version is the ability to link image dimensions to your document's dimensions and aspect ratio. The maximum is 1024x1024 and rounds to the nearest multiple of 64. The dimensions update when you change documents.
More exciting features seem to be ahead. "I'm continuing to look for creative ways to script Photoshop and get results out of SD," said Cantrell.
Get the plug-in here, check out more updates on Twitter and don't forget to join our Reddit page and our Telegram channel, follow us on Instagram and Twitter, where we share breakdowns, the latest news, awesome artworks, and more.