Whether you've decided to follow the path of Stable Diffusion, Midjourney, or DALL-E, generating art with artificial intelligence requires prompts, and experimentation is constant. The so-called “prompt engineering” is real, and it can become intimidating in record time (especially if you don't master English), but the most effective way to start is by using other prompts as a starting point. That is exactly what the Public Prompts portal offers, with examples ranging from sports logos and low-poly animals to everyday objects and anime portraits.
A recurring complaint when creating art with artificial intelligence is that users fail to get precise results. While some elements are harder to generate than others (for example, hands are the nemesis of Stable Diffusion), the most likely issue is a problem with the prompt. As if that weren't enough, it's very rare to find perfect images on the first try, and it's recommended to use a previous generation as a “base” to correct errors and remove superficial details.
So, what about prompts? Our main suggestion remains to visit the Lexica portal, search there for what you want, and copy the prompt from the most appropriate images. However, Lexica is just one option among many available. A user known as “The Ai Art Addict” on Twitter decided to join this trend and create a page called Public Prompts, where they share free, high-quality prompts for Stable Diffusion.
Try the best free prompts to generate art with Public Prompts
In essence, Public Prompts emerges as a response to the new practice of selling prompts, something that “The Ai Art Addict” considers ridiculous (and I share that opinion). For the moment, the page has ten categories: Rendered objects, Funko Pop, Low Poly, Isometric rooms, Surreal microworlds, Adorable creatures, Golden earrings, Sports logos, Block structures, and Hyperrealistic anime portraits.
When we enter a category, we not only see some sample images: We also receive suggestions for the sampler, the number of steps, and the guidance scale (CFG). Each prompt has a highlighted word, and that is the variable to modify. In any case, nothing is mandatory here, and you can play with the prompt by changing its order or combining expressions from elsewhere. The sky is the limit.
To conclude, I must admit that the prompts from “The Ai Art Addict” work much better than I expected. In some cases perseverance is key (say, anime portraits), but overall, we can get positive results with very few attempts.
Official site: Click here
