Latent Consistency Models (LCMs) to overcome the slow iterative sampling process of Latent Diffusion models (LDMs), enabling fast inference with minimal steps on any pre-trained LDMs (e.g Stable Diffusion).

This means you can generate high quality 768 x 768 resolution images in 2~4 steps or even one step , significantly accelerating text-to-image generation. SIGNIFICANTLY reducing inference time and making real time image generation possible.

Signup for the alpha here: