Midjourney VS Leonardo: Full REVIEW
We’ll be taking these two AI- art generation tools and looking at how they fair on a number of different factors.
We’ll start off with the user experience and the user interface, looking at the features and pricing, before putting them to the ultimate duel.
Then, we’ll look at the same prompt in both platforms and see how it compares across a number of different situations.
We’ll take a look at everything from portraits to landscapes, from game asset design to pet photography.
By the end of this video, you’ll have a greater understanding of how these two tools compare.
You’ll understand the strengths and the weaknesses of both.
And most importantly, you’ll know which one is right for your project.
So, let’s dive into our Midjourney vs Leonardo review.
UI Experience and Interface
The first thing to note in our Midjourney vs Leonardo review is that Leonardo is built on top of Stable diffusion which is an open-source AI art generator.
It is essentially a well-rounded UI for Stable Diffusion while Mid-journey is built on its own unique algorithm.
Access to the App
One of the big differences between these two platforms is how you access them.
With Mid-journey, you access it via Discord.
The whole interface is within this chat application.
This brings with it its own benefits and issues.
You type in your command and it comes up in a chat feed.
Compare that to Leonardo which is not built on Discord.
It has its own visual interface.
You essentially have all of your generations built scrolling backward, as well as a visual interface for changing the parameters of your work.
So with this, it’s easy to change the number of images, the dimensions, the guidance scale, and other parameters within your work.
Whereas with Mid-journey, you have to use the UI that exists within that.
So, you have to type in the settings, and from here, you can adjust the parameters.
You do this as well with the text commands, to adjust the aspect ratio or other features like style.
So in that regard, the interface of Leonardo is much more visual and more intuitive.
Since Mid-journey is linked to Discord, it’s going to have a little bit of difficulty expanding its feature set without being cumbersome.
I’d be interested to know if at some point Mid-journey will de-link from Discord.
Viewing Your Generations
The other issue with the Mid-journey user interface is that to view your work,you have to go to a separate app which is a Mid-journey app.
This is where you can view and scroll through all of your creations.
Whereas with Leonardo, they are all within the same platform making it a more coherent user experience.
So, one of the interesting things that Leonardo and Mid-journey offer is the ability to run different algorithms.
In Mid-journey, you can go into settings and change the algorithm.
You can see that there are various iterations of the MJ algorithm as well as an anime-specific algorithm for creating anime-illustrated work.
There are also some older algorithms created focusing on photorealistic images.
There are five or six different algorithms there to use.
However, with Leonardo, there are a lot of different models that you can use.
They have their own Leonardo diffusion model which is one that I’ve found very good for high-contrast images as well as some very specific models.
For example, this origami paper art style algorithm creates some really beautiful work in this very unique narrow context.
There is a pixel art algorithm specifically aimed at creating pixel avatars.
Creating Your Own Models
This brings me to the other thing that’s possible with Leonardo which is to create your own models.
You can upload a certain set of images and allow Leonardo to train itself and create your own fine-tuned models.
This is a bit like using Dreambooth in Stable Diffusion but it makes it a lot easier.
Midjourney currently does not have the opportunity to train your own models.
But with Leonardo, there is an opportunity to explore all of the models that have been created in the community as well.
So, there’s a whole host of opportunities here to refine your work.
Leonardo also has a couple of other features that are built directly into the platform which gives it a little bit more capabilities.
For example, you can instantly remove the backgrounds of images if you’re looking to create a game asset, an icon, or a sticker.
You simply go onto an image that will be suitable.
Let’s take some of these Sprite designs.
Click on Remove Background.
So you click on one of these Sprite images and click on Remove Background.
You can see that here, it has immediately removed the background from our image which is a great feature for speeding up your process.
There’s also an interesting prompt generation tool inside of Leonardo where it will help you build out a more complete prompt.
For example, you can put in a put in a base term like Fantasy Landscape and it will expand on this giving you more ideas of how you can create.
So all you have to do is enter your base term into where it says, Enter a Basic Prompt Idea, and ideate.
Immediately, it would expand on its concept and give you more detailed intricate prompts that you can use.
Now, we have an enchanted forest filled with mysterious creatures and magical plants bathed in soft ethereal light.
So, let’s take this prompt and try it out to go through the actual experience of creating images on both of these platforms.
Then we’re going to take this prompt and paste it in at the same time.
At the same time, we’re going to run this in Midjourney.
So you see that it has taken about 40 seconds to get an image from Leonardo.
We’ve got these two options.
And you can go ahead and generate.
Running Concurrent Jobs
One of the other advantages of Leonardo is that you can have concurrent jobs running.
Whereas in Midjourney even with the Pro Plan, you have a limit of three at a time.
Let’s try and break it in this Midjourney vs Leonardo review.
So you can see that you can simply generate a lot of concurrent jobs allowing you to work quickly.
Whereas in Midjourney unless you upgrade to the most expensive plan, you only have three concurrent jobs.
With the most expensive plan, you have 12.
There is talk from Midjourney about expanding this level set.
The speed of generation is relatively similar, though with more detail.
As you add, it takes longer inside Leonardo.
So if you up the resolution in Leonardo, it does take longer at the moment.
Leonardo is offering a free platform where you get 250 credits per day.
This will change and they will be offering a premium option.
But I believe they will always offer a number of free credits that you can use every day.
Whereas Midjourney has three tiers.
You have the 10-a-month, the 30-month, and the 60-a-month Pro Plan.
In the 30 and 60-a-month plans, they allow you unlimited generations. Leonardo is free for now.
Midjourney has a fairly reasonable pricing structure.
So let’s jump in and take a look at some prompts here in our Midjourney vs Leonardo review.
Woman with a Laptop
The first one we’re going to look at is a female working on a laptop, at the top of the mountain, alpine village, graphic, pink, blue, grain, printed.
So, this on the left is the Midjourney version.
You can see that it’s got a very coherent, consistent image with a nice graphical aesthetic coming out.
Whereas with the Leonardo option, you can see that it has a strong striking contrast and a relatively coherent image.
But overall the consistency is not as intriguing as Midjourney.
There is a little bit of difficulty maintaining coherence in the image of Leonardo with some elongated parts of the body here.
These huge shoulders are on this girl, and here this girl has laptop legs as well like a strange baby-like growth coming out of her head.
But upscaling one of these images improves the coherence a lot.
You can see that this was my favorite.
I upscaled it and it’s a pretty good output.
It could definitely work well for posters.
Whereas comparing this to the upscaled output that I got from Midjourney, there’s a much more refined aesthetic.
The consistency is very good apart from these claw-like hands.
So let’s move on to a mystical landscape in a Studio Ghibli style in our Midjourney vs Leonardo review.
This on the left is the Leonardo output.
I chose this one on the left to upscale and you can say that we’ve got a fairly decent image.
Everything is working well apart from the characters, you could say are letting it down slightly.
But the colors, the composition, and the coherency are fairly good in this instance.
Now let’s look at the Midjourney option which I would have to say is more detailed and more closely following the exact style of a Studio Ghibli film.
But I was using the basic Leonardo diffusion algorithm here.
And I think if you were to train your own algorithm with a set of Studio Ghibli images, you’d be able to get something much closer.
It’s important to note that also these come out differently.
Fine Tuning Prompts
So to really get the best out of each one, you have to wrangle them slightly differently.
As this is built on Stable Diffusion, Leonardo does often require some negative prompts to improve the images.
Whereas with Midjourney, you can pretty much put anything in and the results you get out are beautiful.
So now let’s take a look at comparing some prompts inside of Midjourney and Leonardo.
It’s important to bear in mind that you do need to write your prompts differently to get the most out of these different platforms.
Leonardo is built on Stable Diffusion and it is important to use negative prompts.
Whereas in Midjourney, you can pretty much put anything in and you’re likely to get a fairly aesthetic image out.
So, I’ll show you a few more examples.
There is a chap, hopefully, I’m not butchering your name, who has done some great comparisons between the two.
I will link to his blog post in the description, so you can check out all of his examples.
But here, he was trying out a beautiful landscape of an alien planet.
Comparing Midjourney and Leonardo, you can see,I would say, that the winner for me personally is Midjourney.
But Leonardo is very close.
It has a very different aesthetic.
It’s more constrained and graphic whereas Midjourney is more detailed.
Let’s take a look at some portraits.
On the left is Leonardo with very hard harsh contrasts.
Whereas on the right, we have Midjourney which has a softer more balanced feel.
Both of these create very beautiful life-like women.
Let’s take a look at a pet portrait.
I would have to say that the fur in particular is being rendered more effectively in Midjourney on the left.
And on the right, we do have a beautiful sincere looking cat.
But it’s almost like taking two photographs, two photographers, and asking them to take a picture of the same scene.
The one on the left has taken a little bit more care of the nuance.
It’s almost like there is a slightly more experienced curator of imagery.
Whereas the one on the right is like it’s asking an amateur to take a photograph with an iPhone.
That’s exactly what I think is one of the key differences between Leonardo and Midjourney.
Differences in Prompt Crafting
With Leonardo, you have to be more specific, and more intentional with what it is that you want to get out.
And it takes a little bit more crafting.
But this is like giving you a wider set of brushes, a wider set of tools to create your images with.
It takes some care and craft to get the most out of it.
But if you’re looking to have complete control over what it is that you’re getting out, Leonardo stands out.
Here we have an old-fashioned car and I would say that, in this instance, there is not much to choose between the two.
Most especially if you’re looking at the aesthetic of the image.
Well, actually here, Leonardo on the left has some great evening light.
Finally, we have an old Viking here and I would say that in this instance, both platforms perform extremely well.
These very intense elderly gentlemen have battle-worn eyes.
I always say for me personally, there’s a little bit more life in the eyes of the Midjourney version on the right.
I would also say there’s a slight bit of an issue with the rendering of the fine hairs in his beard here.
In our Midjourney vs Leonardo review, we see that both are extremely capable tools for creating AI art.
I think they have their own strengths and are suitable for use in different use cases.
I like to use Midjourney when I want something quick and to look very aesthetic.
It creates a better coherent image.
But you have the issue that there is not much control that you get inside Midjourney.
You’re essentially quite stuck with what you get out and you don’t have the option to refine or improve it in the way that you are able to do with Leonardo.
Leonardo also gives you the option to build out your own models and use a whole host of additional features.
What is good though, is what I like to do, is to take an image of Midjourney and put it in the Leonardo to develop it further.
Leonardo also has the option of outpainting.
Mid-journey doesn’t have.
So there are strengths to using Midjourney for getting greater aesthetic quality and more detailed and coherent images.
Whereas Leonardo gives you the option of using more features, more capabilities, and more control.
I think there is a place for both of these platforms in any artist’s workload.
I can’t wait to create more with them.
For more on AI, watch this on the left. Let me know what you think of these two AI art tools.
Thanks for watching.
If you liked this article on Midjourney vs Leonardo review, you will also like the articles below.