What Is Midjourney – Everything You Need to Know


If you click on a link and make a purchase, I may receive a small commission. As an Amazon affiliate partner, I may earn from qualifying purchases.
Read our disclosure.

Key Takeaways

  • Midjourney is one of the best and first text-to-image AI art generators.
  • Midjourney was founded in 2021 by entrepreneur David Holz in San Francisco.
  • There are already over 13,000,000 Discord members on their server.
  • Pricing starts at $10/month.
  • Midjourney is constantly developed, with new features added to the AI art generator almost monthly.
  • Midjourney offers two modes, “fast” and “relax,” for image generation.

Introduction to Midjourney


Midjourney is one of the best and first text-to-image AI art generators, with a community of over 13,000,000 Discords members. Midjourney was founded by David Holz in 2021 in San Francisco, United States. AI art generators such as Midjourney can convert text-to-image with a simple text prompt such as “mad max style real people riding 20 post apocalyptics cars and trucks” (below images).

Midjourney can create amazing images with a simple text prompt.

Midjourney can create photo-realistic images, anime, cartoons, isometric images, etc. The results depend on the text prompt you use. The more detailed, intricate, and comprehensive the text prompt, the better images you can expect.

Every time you use the generator, you get four unique and different images in a 2×2 grid (with a 1:1 aspect ratio (default)), from which you can create variations or upscale the one you like the most.

Midjourney works by using large language and diffusion models to create unique images based on text prompts.

Text prompt example:

cybergoth James jean, Kuvshinov Ilya, moebius, nujabes art, Watercolored pencil :: no borders :: --ar 2:3 --v 5 --s 750

Example results using text prompts on Midjourney.

About Midjourney

Midjourney is an independent research lab on a quest to discover unexplored realms of thought with a vision of expanding the imaginative capabilities of the human species. The team at Midjourney consists of a small, self-funded group focused on design, human infrastructure, and AI.

Their primary objective is to delve into new frontiers, exploring unconventional mediums and pushing the limits of human creativity.


Midjourney’s pricing.

With Midjourney, you can select monthly or yearly pricing (with a slight price reduction). The Basic plan ($10/month) does not include the “Fast” image generation mode but otherwise is the same as the Standard plan.

The Pro plan ($60/month) includes 15h more “Fast” image generations (compared to the Standard plan ($30/month)) and stealth image generation, otherwise being the same as other plans.

The Mega plan is the latest pricing plan and offers a whopping 60 hr/month of Fast GPU time. The pricing for this plan starts at $120/month, yearly subscription being a bit more affordable.

How many images can you generate per plan?

On average, you can generate 60 images with one GPU-hour.

  • Free plan: relax mode and if availability (highly unlikely).
  • Basic plan: ~200 images/month (relax mode).
  • Standard plan: ~900 “fast” images and unlimited “slow” images.
  • Pro plan: ~1,800 “fast” images and unlimited “slow” images.
  • Mega plan: ~3,600 “fast” images and unlimited “slow” images.

Midjourney Versions

Image showing the capabilities of Midjourney. Image credits.

Midjourney regularly introduces new model versions to enhance efficiency, coherency, and image quality. Users can switch between different models by adding the --version or --v parameter to their text prompt or by using the /settings command and selecting a specific model version.

Each model has its strengths and specializes in producing different types of images.

Model Version 5.2

V5.2 brings several exciting new features and enhancements to Midjourney. One of the notable improvements is the New Aesthetic System, which results in sharper and improved aesthetics, along with slightly better coherence and text understanding. Additionally, diversity has been increased, though you may occasionally need to roll more than once to get your desired outcome.

A significant addition is the “High Variation Mode,” now turned on by default. This mode significantly increases the variety of generated images for all variation jobs. Users can toggle this feature by typing /settings and selecting a different variation mode. Moreover, users now have the option to choose the strength of variation they want under all upscales.

The introduction of the /shorten command allows users to analyze a prompt and receive suggestions on redundant or unnecessary words, helping to refine and optimize prompts for better results.

Another exciting feature is the “Zoom Out” option, available for all upscales. Users can find ‘zoom out’ buttons underneath each upscale to reframe the image. The [Zoom Out 1.5x] and [Zoom Out 2x] options pull the camera out and add details on all sides.

[Make Square] is useful for transforming non-square images into square ones by adding details on two sides. For advanced users, [Custom Zoom] provides a popup text box to change the prompt and aspect ratio as you zoom out, eliminating black bars from previous images.

Currently, V5.2 is enabled by default to maximize testing and feedback, but users can easily switch back to V5.1 by typing /settings and clicking V5.1. The High Variation Mode is exclusive to V5.2, while the /Shorten command does not work with multi-prompts.

Panning in Midjourney offers a fascinating way to expand and extend your images (similar to outpainting in Stable Diffusion). Below the upscales, you’ll find arrow buttons ◀️ ▶️ 🔽 🔼 that enable you to extend the image in different directions.

By clicking on an arrow, the image will expand accordingly. For more creative possibilities, you can activate Remix mode by typing /settings and clicking on it. This allows you to change your prompt as you extend the image, making it perfect for telling panoramic stories.

Interestingly, as you pan out the image, the generator will keep extending it, even though the resolution becomes considerably large. This exciting feature means you can pan multiple times, giving you endless opportunities for artistic exploration.

Model Version 5.1

The Midjourney V5.1 offers a stronger default aesthetic, making it more user-friendly with simple text prompts. It excels in coherency, accurately interprets natural language prompts, produces fewer unwanted artifacts and borders, and provides sharper images.

It also supports advanced features like repeating patterns using the --tile command.

Model Version 5 (Midjourney V5) generates images that closely match the given prompts, particularly focusing on photographic generations. However, this model might need longer prompts to achieve the desired aesthetic.

Model Version 4

Model Version 4, Midjourney V4, was the default model from November 2022 to May 2023. It introduced an entirely new codebase, and AI architecture developed by Midjourney and trained on the Midjourney AI supercluster.

Unlike earlier versions, Model Version 4 possessed enhanced knowledge of creatures, places, and objects and exhibited high coherency, making it proficient with image prompts.

Niji Model 5


The Niji Model Version 5 is a collaborative effort between Midjourney and Spellbrush, specializing in producing anime and illustrative styles. It offers extensive knowledge of anime, anime styles, and anime aesthetics, excelling in dynamic/action shots and character-focused compositions.

Users can fine-tune this model using --style parameters such as cute, scenic, or expressive (for example, --style expressive) to achieve unique looks.

Switching Models

To switch between models, users can either add the corresponding version parameter (e.g., --v 4, --v 5, --v 5.1) or use the /settings command and select the desired version from the menu.

Future of Midjourney Models

Midjourney has evolved from versions 1, 2, and 3, each version bringing new and improved image quality to the art generator.

Midjourney has stated that they are continuously improving their AI art generator. While the user interface is still clunky (thanks to Discord), it’s only expected that it will be transferred from Discord to the Midjourney app at some point (already in discussion).

Midjurney has also hinted that they will be creating an app for Midjourney, one that enables you to create images.

We also expect to see one version of Midjourney to be used so that users don’t have to switch between versions to achieve a certain image quality and aesthetic.

As there are already many AI art generator alternatives to Midjourney, Midjourney must continuously evolve its AI art models to enhance the image generation experience, providing users with various options to suit their preferences and needs.

Feature image credits.



Digital Artist

I’m a digital artist who is passionate about anime and manga art. My true artist journey pretty much started with CTRL+Z. When I experienced that and the limitless color choices and the number of tools I could use with art software, I was sold. Drawing digital anime art is the thing that makes me happy among eating cheeseburgers in between veggie meals.

More Posts

Thank You!

Thank you for visiting the page! If you want to build your next creative product business, I suggest you check out Kittl!