Friday, September 30, 2022

5 Best AI Art and Image Generators In 2022

Artificial Intelligence is rising and thriving in 2022. From creating genuine music to art, it is developing at a rapid pace and may soon change how most businesses function for either better or worse.

Here I am going share some amazing AI tools that can create genuine art and images based on your text prompts or from referencing images you upload.

The best part about all of this is it happens in a matter of seconds, creates genuine art and images that can be used without worrying about copyright!

Top 5 AI Art And Image Generators

Over the last couple of months I have extensively spent my time searching for the best and most promising AI art and image generators.

The following 5 tools are the best I found, but there are still hidden gems out there that I may not have come across.

If you know any better or equivalent AI other than those mentioned here, do mention it in the comment section.

1) Midjourney

Midjourney was the first AI art generator that I came across around 5-6 months back that obliterated my outlook of AI generated art.

hell-and-heaven-midjourney

Before Midjourney, the AI art or image generators I came across where pretty underwhelming.

Seeing the true potential of what AI can do really spiked my interest.

Midjourney is an independent research lab, self-funded with a fully distributed team.

For the moment, this AI can be accessed from their Discord channel where new users get limited credits, around 25 queries in total.

They also have memberships with the Basic Membership of $10 USD which allows 200 images per month and the Standard Membership of $30 USD with unlimited personal use.

Now, how does it work?

Midjourney is a text to image creation tool, so all you need to do is use the /imagine prompt and enter the query you want like ‘Monkey playing football’ and it will generate 4 images related to the prompt.

monkey-playing-football-midjourney-discord

Here is a better look at the generated images:

monkey-playing-football-midjourney

You can now get more variations or upscale any of these images you like, or you can generate entirely new set of images.

U is for upscaling and V is for Variations.

The number beside is for the image you want to change. The redo sign is for getting a completely new set of images.

This is how the upscaled version of the 3rd image looks:

monkey-playing-football-midjourney-single-upscaled-image

Now this is not everything. You can stack queries one on top of another to generate an even more unique image.

This way you can add elements for the image and even the image style.

So, let’s try something a bit complex like ‘monkey playing football, heaven, angels watching as audience’.

This is what I got and I am very much satisfied (and slightly amused) with the results.

monkey-playing-football-heaven-angels-watching-as-audience-midjourney

I only used these weird text prompts to show how much randomness Midjourney can easily handle.

You can create countless combination of images with stacking different queries. Like this is what ‘Firestorm in Space’ was generated.

midjourney-firestorm

Other than that, you can also add reference images while adding the prompts.

You can even add specific prompts to remove certain elements, like ‘–no trees’ to not add any trees. Here is a complete guide for all the commands for Midjourney.

Your imagination is basically the limit. I have seen users write 100-150 word prompts, creating highly detailed images.

Visit the Midjourney community showcase and have a look at the work of others.

The images showcased in the gallery also reveals the text prompts used to generate it, giving the other users a better idea on how to create it.

Other than that, there is the Discord channel that you can enter and see users generating images live while also having a go at your own creation.

2) DALL-E 2

DALL-E 2 is probably the most popular one on this list as it recently had a huge media break.

The successor of DALL-E is a more powerful and highly accurate with its image creations.

Sadly, I am still on the waiting list for it and haven’t had the privilege of using it first hand.

However, I have seen others using it and it is easily one of the best AI image generator out there at the moment.

I will update this post once I get the access and give a more detailed review on Dall-E 2.

Dall-E 2 can create original, realistic images and art from the text you input.

Like Midjourney, it also understands different elements and combines multiple attributes, styles and concepts to form a unique image as prompted from the text.

astronaught-riding-a-horse-dall-e-2

Another unique feature of DALL-E 2 is that it can make realistic edits to existing images from a language caption and similarly add and remove elements like shadows, reflections and textures.

dall-e-2-add-element-to-image

Other than that, DALL-E 2 can use an image and create different variations for it, generating completely unique images that can be used copyright free.

create-image-variation-Dall-e-2

The only downside of DALL-E 2 at the moment is it’s not yet accessible to all.

New users have to enter their waiting list, which is quite a bummer. But they are still in the Beta phase and likely not yet prepared for mass use.

3) NightCafe

NightCafe is another amazing image and art generator. It may not be near the level of Midjourney and Dall-E 2 but is a lot more accessible.

First you will have to sign-up and then click on the Create option. Here you will see two options, Text to image/video and style transfer.

The text to image/video will generate an image or video based on your text prompts and the Style transfer changes the style of the image you upload to the style you select.

Changing the style of the image is simple. First you upload an image, like let’s say an image of the Statue of Liberty.

input-image-for-change-night-cafe

Next select the style that will be merged with the image.

style-of-the-image-night-cafe

Set the resolution of the image and generate it. This is what I got:

changed-style-of-statue-of-liberty-from-night-cafe

Exciting isn’t it? And I didn’t had to pay anything because of the free credit system of Night Café, (which I have explained later on).

Next, I explored the Text to image generator and found it pretty good as well. Although the results weren’t as good as Midjourney, it was still amazing for an AI standard.

First you have to enter the text prompt and then choose the style, like ‘old man playing football on the moon’ with the ‘sinister’ style.

night-cafe-choose-style

Next you have to choose the Algorithm that you want to use. Both Algorithms have their strengths and weaknesses on art generation.

night-cafe-algorithm-for-image-creation

After completing the steps, this is what I got:

old-man-in-night-cafe

Now, it definitely missed some prompts but the end product is amazing nonetheless.

So, how do you get access to NightCafe? NightCafe runs on a credit system. Various tasks like creating an image or enhancing it consumes credits.

These credits can be purchased or earned by being socially active.

There are badges or achievements that give free credits, like 1 credit for liking 5 posts in their community gallery. Also, they give 5 free credits every day.

4) Artbreeder

Artbreeder is a very unique art or image generator that is completely free to use.

It moves away from the text to image creation method and instead focuses on mixing different images and creating something completely new.

This is also the best place for generating Characters or faces with its unique gene system.

So, here is how it works:

First you will have to click on the plus icon (+) icon to start the creation process.

Next you will have to choose the image type you want to mix.

create-image-categories-artbreeder

You can also upload your own image but that limits the AI to Portraits and Landscapes.

create-upload-categories-artbreeder

So, let’s select an image and create a portrait. This is where I think Artbreeder really shines.

Here you will see some auto generated images and a space to add a ‘Parent’.

create-portrait-art-breeder

You will have to add parents, which are basically images, or portraits in this case, which you want to mix.

After you have selected the images you want to mix, Artbreeder will automatically mix the images and generate portraits based on all the characteristics.

mixing-multiple-images-to-create-portrait-in-artbreeder

Now you can tweak it further by deciding which face and style from the images you want to be dominant in the new portrait. The images will change based on the tweaks you make.

Remember that these are randomly generated so it won’t generate the same image each time, even if you input the same tweaks.

Tweaking-the-portraits-in-artbreeder

Also, you can further tweak these images from the right menu, which will generate the prompts you enter even if they don’t exist among the parent images. Like changing the gender or adding a fully grown beard.

Next, click on the images you like to save them. Once you save an image, it will be separated and another new render will take its place.

The saved images will also appear on your Profile.

save-images-art-breeder

What’s more interesting is that the images you have saved can be remixed further with other images or parents and this can go on and on.

You can even remix an image created by other users, changing them as you like.

Another great feature is the image history, which basically shows the entire ‘lineage’ of the image and how it came to be!

image-history-art-breeder

Similarly, you can also combine different images and create landscapes or sci-fi arts. The possibilities are limitless!

artbreeder-creation

Artbreeder is completely free to use. They also have premium subscriptions, which provides certain privileges like being able to download the images at a higher quality and is basically a way to support the team of Artbreeder.

Definitely subscribe if you like using Artbreeder as it will support the devs.

5) Starry AI

Starry AI is another top AI art generator with an easy to navigate UI. It is also a text prompt to image creation AI.

Their AI is quite advanced and is capable of creating pretty realistic images with different styles and complicated text prompts.

Like Night Café, Starry AI also provides 2 AI algorithms for image creation, Altair and Orion.

select-ai-in-starryai-for-image-creation

After tinkling with both, I found Altair to be better with scenery images while Orion is better with character focused images.

After choosing the Algorithm, you will have to enter the Text prompt.

starry-ai-image-creation-prompt

Next you will have to add a style. You can pick multiple styles and not just one.

This is where Midjourney is better as it is not limited to a set number of styles.

starryai-style-selection

After picking the style, you can change the advanced settings. These settings include setting the aspect ratio of the image, changing the model type, and setting the Runtime.

The more Runtime you will set, the more time the AI will have for creating the image leading to a better result.

starryai-advanced-settings

After doing all that, you will have to click on Create to produce the image. It will take around 2 minutes to form the image.

flying-man-in-space-finished-starryai

Now you can do a few things here. The Share and Download tabs are pretty self-explanatory.

The Generate Animation option creates a video of the image formation from the first iteration to the end and is a nice addition.

The Upscale option upscales the image quality to what you want. Here is how the 8x upscale of the above image looks like:

flying-man-in-space-starry-ai

Next is the Evolve option, which basically adds the created image as the reference image in the creation menu and lets you change the initial text prompt and create a more specific image.

evolve-created-image-starryai

Starry AI has a similar credit system to NightCafe. Each tasks that you perform requires credits, which can be purchased.

These credits can also be earned by sharing your creation on the social media. Also, you get 5 free credits every day!

Wrapping Up

So, that’s it! AI is developing at an unimaginable speed and businesses and lifestyles will also have to change accordingly to survive.

Remember to comment about any AI you think should be on this list and help others have a look at it.

Sanmay Chakrabarti
Sanmay Chakrabarti
Sanmay is a Windows Insider and editor at BottoBotto. He is a Tech enthusiast and has been writing tech blogs for over 2 years now. He loves reading books, traveling to new places and listening to music in his free time.

Related Articles

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Latest Articles