Stable Diffusion

Stable Diffusion is a source-available, deep learning text-to-image model.

Model version history
Stable Diffusion model software was initially had been released in August 22, 2022.

The model called depth2img was released in November 24, 2022, and so version 2.0 of Stable Diffusion.

Generating images
True to it's purpose, Stable Diffusion generates detailed images conditioned on text descriptions. It can also be applied to other tasks such as inpainting, outpainting, and generating image-to-image translations guided by a text prompt.

Models may fail to natively generate anything above 768x768 or below 320x320, with pre-2.0 models troubling to fluently generate images at 512×512 resolution. As a results, generated images noticeably degrades.

Self-hosting
Stable Diffusion's code and model weights have been released publicly, it can be found in GitHub and Hugging Face. That means that Stable Diffusion can be self-hosted. It is not recommended to do that for computers with less than 8 GB VRAM.