Have you heard about the award-winning paintings created by artificial intelligence? Yes, do not be surprised, this is what happened. There are many sites that provide services to create semi-real images using artificial intelligence based on text descriptions, and it depends on conversion models that began to be developed in 2010 due to the great progress in the field of neural networks.
These models accept a description in natural language as input and produce an image Match the description as closely as possible. Including the Stable Diffusion model, which was released in 2022. It is open source and has been trained on billions of images. It has the ability to produce results commensurate with the quality of real images and art drawn by humans. In this post, we will learn about a group of the best sites that use this model wonderfully.
- Stable Diffusion Web
This website uses the Stable Diffusion model to produce near photorealistic images from purely text input. Simply write a few words about the image you wish to create using natural language and then click Create Image. It takes some time for Stable Diffusion to process the text and generate two images using its AI model and the size of each image is 512 x 512 pixels.
- NightCafe
Another text to image generator based on Stable Diffusion which can help you to create incredibly aesthetic images from a text prompt and also allows to define the style of the image. Just write a few phrases in English, choose a picture style and click "Create". The text-to-image form may take a few seconds to process the prompt and generate the image.
By clicking on "More Options" you will find additional styles such as Portrait, Black & White, Modern Comedy and more. With Advanced Options enabled, you can choose the aspect ratio, how the image matches the text vector, how many images to create, and more. You can create a maximum of 10 images after which you must purchase additional credits.
- Hugging Face
Another photo maker with text that works like the previous ones and will produce 4 images of 512 x 512 pixels each.
