Stable Diffusion XL (SDXL)

Stable Diffusion XL (SDXL) is a text-to-image model that turns a written prompt into a picture. Compared to earlier Stable Diffusion versions, it produces higher resolution, more detailed results and follows prompts more accurately, while still being efficient enough to run on consumer GPUs.

In tendre.red, SDXL checkpoints such as CyberRealisticXL, RealVisXL, JuggernautXL and DreamShaperXL power text-to-image generation locally on your own GPU, and the same checkpoint is what your LoRAs are trained against.

Related tutorial