Skip to main content
HomeTutorialsArtificial Intelligence (AI)

How to Use Midjourney: A Comprehensive Guide to AI-Generated Artwork Creation

Discover the power of Midjourney, a generative AI tool for creating stunning artwork. Learn how to get started, write effective prompts, and optimize your usage with our step-by-step guide.
Jul 2023  · 12 min read

You don’t necessarily need to be a talented artist to create compelling imagery anymore. With just a few prompts, you can generate interesting pieces in a matter of minutes. The secret is Midjourney.

Midjourney leverages the creativity of generative AI to do much of the heavy lifting. Your job is merely to pass it a little bit of text to describe what sort of image you’re thinking of, and it creates it.

Despite being a more recent development, Midjourney has taken the generative AI space by storm. It’s earned itself a spot alongside some of the best AI systems out there, like DALL-E and Stable Diffusion.

So let’s see how you can get started with Mindjourney to generate your own images.

How to Get Started With Midjourney in 4 Simple Steps

Step #1: Sign up to Discord

If you don’t have a Discord account, you will need one – register to create a free account. You can access the Midjourney Bot can from anywhere Discord can be used (e.g., web, mobile, desktop applications).

Step #2: Sign up for Midjourney

Once you’ve got Discord up and running, head over to the Midjourney website and select “Join the Beta” – or you can click on the Discord invite link to go straight there. The next page will be an invitation to the Midjourney Discord channel. When you accept, you will be granted access to the Midjourney server.

The Midjourney Discord invite

The Midjourney Discord invite

Step #3: Pick your Midjourney plan

There are three subscription tiers for Midjourney. Each one grants you access to the Midjourney member gallery, the official discord server, general commercial usage terms, and more – see the Midjourney documentation to get a full list of the price plans. To subscribe to a plan, use the /subscribe command in one of the newcomer rooms (e.g., #newbies-29) to generate a personal link to the subscription page.

Midjourney plan comparison - Image Source

Midjourney plan comparison - Image Source

Step #4: Start generating images

If you’re not ready to subscribe just yet, you can start creating prompts from the newbie channels – more on that in the section below.

Note: the newbie channels can quickly become overwhelming due to several requests being posted at any one time.

Being a paid user allows you to send prompts directly to the Midjourney Discord bot. However, they will still be publicly visible in the member's gallery; If you want to generate private images (stealth image generation), you must subscribe to the Pro Plan.

How to Generate Your First Midjourney Image

To get started generating your first image, go to one of the newcomer channels in the Midjourney Discord server; they’re typically named “#newbies-” followed by a number. Don’t get too hung up on which one to choose – there’s no real difference between them.

The way to start your prompt is by using Discord commands. Simply enter “/” followed by “imagine” and the prompt you want Midjourney to create.

Here’s an example:

/imagine prompt: musician looking at his laptop excitedly

Image from Discord Server

Image from Discord Server

It takes about a minute to generate a response, but when it does, the Midjourney bot will render four different images based on your prompt by default.

Note: Midjourney no longer offers free usage except during some brief promotional periods. If you would like to send a prompt, you must subscribe.

Editing and Refining Your Midjourney Images

Once Midjourney has finished generating your image, the bot will return some prompts for you to edit and refine your images. Here are a few options you have:

U buttons

These range from U1-U4. We use U buttons to upscale an image. This generates a larger version of the selected image as well as more details.

V buttons

These also range from V1-v4. You can use V buttons to create slight variations of the selected grid image. A new picture grid created as a variation is modeled after the general look and composition of the selected image.

The 🔄 (re-roll)

There’s also the option to rerun the job; this will rerun the original prompt and produce a new grid of images.

How to improve your Midjourney prompts

Options that are added to a prompt as parameters alter how an image generates. The Aspect Ratios of an image, the Midjourney Model Versions, the Upscaler being utilized, and many other things can all be changed via parameters.

Parameters are always added to the end of a prompt. You can add multiple parameters to each prompt.

/imagine prompt: musician looking at his laptop excitedly --aspect 2:3 --stop 95 --add birds

Advanced Midjourney Features

Images may also be included in a prompt to alter the composition, design, and color scheme of a Job. This means you can utilize text prompts and image prompts, or you could use either method independently.

To get started with an image prompt, enter or paste the image's website address after the /imagine Discord command. Note that the web address must end with a .png, .gif, or .jpg extension to work.

If you wish to add any additional text and parameters to complete the prompt, it must be done after adding the image address. For example:

/imagine prompt: http://imageURL1.png description of what to imagine --parameter1 --parameter2

To recap, here are the rules of using image prompts:

  • Prompts with images must start with the image web address
  • For prompts to function, there must be two images or one image and text
  • A direct link to an image on the internet is required for an image URL
  • Your file's extension should be either .png, .gif, .webp, .jpg, or .jpeg
  • To obtain an image URL in most browsers, you must right-click an image and select ‘Copy Image Address’
  • A streamlined image prompting procedure designed for mobile users is the /blend command

Uploading images from your computer or phone

We’ve already covered that an image address link ending in .png, .gif, .webp, .jpg, or .jpeg is required if you wish to include an image in your prompt, but what if the image is on your computer or phone?

In that case, you would need to send the image to the Midjourney bot first to generate a link.

10 Tips On How to Write Effective Midjourney Prompts

If you’re ready to get started creating images with Midjourney, these top tips can help you get the most out of the generative AI tool:

1. Be concise

Midjourney generates the best results when your prompts are specific. Include as many specific details as possible in your descriptive language. However, keep your prompts concise. Anything written after approximately 40 words is likely to be ignored; after 60 words, it becomes highly likely to be ignored, and after 80 words, it will almost certainly be cut off.

2. Use descriptive language

Use adjectives and adverbs to add nuance and emotion to your prompts. They can describe the color, shape, size, texture, and other characteristics of the object or scene that you want to create.

3. Avoid ambiguity

Use clear and direct language that accurately describes the image you want to create. Ambiguous language can be interpreted in a variety of ways, which may lead to unexpected results.

4. Experiment

You must experiment with the model to understand how it interprets various keywords. Check the outputs you’re getting and see how small adjustments to your prompts and settings can yield different results.

5. Use a prompt generator

Prompt generators are built for the sole purpose of assisting you to write descriptive Midjourney prompts. Using them can give you some inspiration and serve as a great foundation for your initial prompt.

6. Use the creative tag

The creative tag will steer Midjourney to generate more unconventional outputs. This is a great way to come up with new ideas or stray from the traditional appearance of Midjourney content.

Here’s an example of how your prompt would look:

/imagine man playing cricket --test --creative

7. Add weights to your prompts

With the double colon(::), you can instruct Midjourney to interpret terms separately. One such place this is handy is when you want to increase or decrease the importance of certain terms in your prompt.

An example taken from the Midjoueney documentation is the term ‘space ship.’ As all words are considered together, the Midjourney Bot produces images of sci-fi spaceships. If the prompt is separated into two parts, space:: ship, then each concept is considered separately, generating a ship sailing through space.

Image source

Image source

8. Use negative prompts

Negative keywords are terms you do not want to appear in your image. Many advanced users add negative prompts like (–-no hands, no text, etc.) to create more nuanced images.

9. Consider subscribing to use private mode

There are various other Midjourney users in the discord, and they’ll be sending their prompts just as you would. This can make it a little overwhelming when you’re trying to work on your own projects. If you want to gain a little more clarity, consider subscribing to get access to private mode.

10. Modify your settings

Enter /settings into Midjourney to bring up the settings screen. From here, you can select all the settings you wish to include in your prompts without having to constantly write them out manually.

Conclusion

Midjourney makes creative artworks accessible to all with a few simple descriptive keywords and/or images. Getting started with the is not as difficult as one would think initially. By following this tutorial, you have enough resources to start creating some interesting projects.

If you want to learn more about Generative AI, you may be interested in the following:

Midjourney FAQs

What is Midjourney?

Midjourney is a text-to-image generator powered by generative AI. It can generate highly realistic images based on the text prompts provided by users.

What are prompts in Midjourney?

Prompts are short text phrases that the Midjourney Bot interprets to produce an image. They can be as simple as a single word, phrase, or emoji. The bot breaks down the words and phrases in a prompt into smaller pieces, called tokens, that can be compared to its training data and then used to generate an image.

What are the types of prompts in Midjourney?

Midjourney offers three types of prompts: text prompts, image prompts, and multi-prompts. Text prompts require you to provide a textual description of the image you want to create. Image prompts involve uploading an image as the input to generate a new image. Multi-prompts combine text and image prompts to generate images that correspond to both types of input.

How can I write effective prompts in Midjourney?

To write effective prompts, be specific and concise, use descriptive language, avoid ambiguity, and experiment with different prompts. For complex images, consider using multi-prompts. You can also use negative keywords to exclude certain elements from your image.

What is the –creative tag in Midjourney?

By adding the –creative tag to your prompt, you can guide Midjourney toward outputting more unique or unusual images. For example, you might notice a more vivid color palette or abstract design.

What is the –stylize tag in Midjourney?

The –stylize, or –s, tag will encourage Midjourney to take additional creative liberty with your prompt. The –stylize value ranges from 0 to 60,000. The higher the stylize value, the more abstract your images will become.

What is the –ar tag in Midjourney?

By adding the –ar tag at the end of your prompt, you can tell Midjourney which image size your outputs should appear in. This is perfect for creative projects or tiling various images together.

What are the subscription plans for Midjourney?

Midjourney offers four subscription tiers: Basic, Standard, Pro, and Mega. Each plan includes access to the Midjourney member gallery, the official Discord, general commercial usage terms, and more.

  • Basic plan. Costs $10 per month or $96 per year ($8 per month). It includes 3.3 hours of Fast GPU Time per month.
  • Standard plan. Costs $30 per month or $288 per year ($24 per month). It includes 15 hours of Fast GPU Time per month and unlimited Relax GPU Time.
  • Pro plan. Costs $60 per month or $576 per year ($48 per month). It includes 30 hours of Fast GPU Time per month and unlimited Relax GPU Time.
  • Mega plan. Costs $120 per month or $1152 per year ($96 per month). It includes 60 hours of Fast GPU Time per month and unlimited Relax GPU Time.

You can purchase additional Fast GPU hours for $4/hr. These purchased hours do not expire, but you must have an active subscription to use them.


Photo of Kurtis Pykes
Author
Kurtis Pykes
Topics
Related

blog

Using Generative AI to Boost Your Creativity

Explore art, music, and literature with the help of generative AI models!

Christine Cepelak

14 min

blog

7 Essential Generative AI Tools for Building Stand-Out AI Applications

Dive into the technical building blocks of Generative AI, from APIs to LLMOps tools for sustainable app development.
Abid Ali Awan's photo

Abid Ali Awan

9 min

blog

What is Multimodal AI?

Discover multimodal AI, one of the most promising trends in generative AI.
Javier Canales Luna's photo

Javier Canales Luna

8 min

blog

What is AI? A Quick-Start Guide For Beginners

Find out what artificial intelligence really is with examples, expert input, and all the tools you need to learn more.
Matt Crabtree's photo

Matt Crabtree

11 min

tutorial

Vertex AI Tutorial: A Comprehensive Guide For Beginners

Master the fundamentals of setting up Vertex AI and performing machine learning workflows.
Bex Tuychiev's photo

Bex Tuychiev

14 min

code-along

A Beginner's Guide to Prompt Engineering with ChatGPT

Explore the power of prompt engineering with ChatGPT.
Adel Nehme's photo

Adel Nehme

See MoreSee More