What's the best open source or pre-trained text to image model?

latent diffusion model trained on LAION-400M https://github.com/CompVis/latent-diffusion