Leonardo.ai is a new image creation tool that comes to specialize in a wide number of topics.
As soon as we register, it asks us for details about what we will use the work for, and offers us several options to customize its result.
Available at leonardo.ai, by invitation, I have already been able to try it, and those are my conclusions.
Advantages of Leonardo.ai
Leonardo.ai has an important advantage, since it allows us to choose the model. We can select a model specialized in creating manga, another more photorealistic, another more artistic… Once the model is selected in the column that you see in the image below, we can also specify the weight of our prompt, in case we want it to follow it to the letter or No.
The additional models to the ones you see in the image above are those of the platform itself and those created by the community. That means that an artist can create a model that suits her style, and thus create new works that have her particular signature.
In the configuration we can decide what type of work we want:
and allows us to create images from other images
As you can see, what really makes Leonardo AI stand out is the ability to train your own models. This is a unique feature that is not found in other similar tools. Images of a similar style can simply be uploaded and saved as a template for future image generations. You can also share the model with the community, which extends the capabilities of the tool and makes it more collaborative.
Can it be compared to Midjourney?
They are different things. Midjourney requires you to specify everything at the prompt, and doesn’t allow you to use other models than yours, but the results are more attractive, there’s no doubt about that. Although Midjourney remains the leading tool when it comes to the quality of generated images, the emergence of Leonardo AI is a breakthrough in the field of artificial intelligence and computer generated images. The ability to train your own models and collaboration with the community is something that really makes it unique.
I used this available Midjourney prompt:
geeky man with glasses full body hyper realistic + cinematic shot + photos taken by ARRI, photos taken by sony, photos taken by canon, photos taken by nikon, photos taken by sony, photos taken by hasselblad + amazingly detailed, sharp, details + professional lighting, photography lighting + 50mm, 80mm, 100m + lightroom gallery + behance photos + unsplash –v 5 –s 750
and compared it to what Leonardo.ai would generate with these characteristics:
As you can see, it is still far from being able to achieve the hyper-realistic results of Midjourney v5, but if we compare it with v4, the difference is not that much:
Conclusion
I see Leonardo AI as a tool for designers, who can create their own unique models to carry out their work in a more controlled way, although we already know that the final use of each tool is not dictated by logic.
