While image diffusion models add continuous noise to pixel values, text diffusion models can't apply continuous noise to discrete tokens (chunks of text data).

Text to image AI makes artwork in seconds. There is tons of free software available, as well as some with subscriptions that can be used to generate artwork.