Artificial intelligence has taken a big leap in recent years. Although few of us thought that AIs were going to replace creative work, now we have found out that not only are they very good at it, but it even makes us ask the dreaded question: will they ever replace the human hand?
As amazing as they may seem, the rise of artificial intelligence has also raised many blisters. This has caused the public to take a close look at how these models work and how they are trained, giving rise to ethical and legal debates.
However, none of this has managed to stop the creation of more and more generative models. The generators of images from text were some of the first tools that managed to capture the attention of the public. Here we leave you a list with the best artificial intelligences to create art from a description.
Like all the AIs on this list, with StarryAI you will be able to generate images very easily. The technology behind this model is quite versatile and offers you a multitude of options: you can choose between different algorithms, image sizes, number of iterations (the more iterations, the better results) and artistic styles.
One of the most remarkable aspects of StarryAI is that it is available as an app for both Android and iOS. You can also use its website, although you should keep in mind that you will have a limited number of images (to have more, you will have to get your wallet out).
Stable Diffusion is one of the most popular names in the world of AIs. Specifically, it is an open source machine learning model developed by Stability AI. It is capable of generating images from a text description and can also modify existing photos or drawings.
It has been trained with billions of images and generates surprising results, to say the least. It was launched on August 22, 2022 and since then, it has not stopped being on everyone’s lips.
Midjourney is a tool focused more on the artistic field and that creates unique images starting from a brief written description. David Holz heads the development team behind Midjourney and is also the co-founder of Leap Motion, a motion sensor with which a computer can be controlled with just our gestures.
More than photos, Midjourney recreates illustrations. Any idea is valid for the tool to generate several samples (and with quite striking colors). Midjourney worked as a beta during 2022, but today it already has a definitive version available to anyone. With a free account you can generate up to 25 images, but if you need more, you must subscribe to their service.
DALL-E 2 is the revision of the artificial intelligence model we saw last year, and it’s more spectacular than ever. OpenAI defines it as “a new AI system that can create realistic images and art from natural language description.” It is currently in a closed testing phase, but there are numerous examples of how it works.
Like the previous model, this AI takes advantage of the potential of GPT-3 to generate its designs, using a complex system in which an image is first generated given a text caption, and then a second image is generated conditional on the first.
NightCafe has one of the most powerful technologies for creating images and stands out for having more computing power and many configuration options.
Like the rest of the models, you will only have to write a description of what you want to see and then choose an artistic style (oil, cartoon, concept art, surrealism, etc.)
According to what we can read on its official website, Craiyon uses a Dall-E mini model (known as Dall-E mega). Boris Dayma, creator of Craiyon, explained that for the creation of Craiyon he based himself on DALL-E, the first version before DALL-E 2 came out.
Crayon is the result of creating a smaller version of the OpenAI model. It is open to the public for free use and is also free. If you wonder why this service is free, Craiyon explains on its website that this is possible thanks to the money provided by the ads and advertising that the tool has.