MidJourney, Ultimate Guide — Part1, Image Generation Basics

BoredGeekSociety
5 min readDec 21, 2022

--

Everything you need to know to leverage MidJourney!

In this series, in three parts, we’ll cover:

(1) Everything you need to know to start using MidJourney,

(2) Then look at how you can optimize your prompts, to build various styles of images and even UI designs for websites.

(3) And finally see how you can rely on NoCode/LowCode automation tools to build a full fledged image generation pipeline with MidJourney.

If you enjoy this kind of posts and would like to see more of them, please consider supporting us: 🔔 clap & follow 🔔

Dall-E 2 and Stable Diffusion are two notable text-to-image models that have gained attention in the field of artificial intelligence and machine learning. Dall-E 2 was the first language model to be able to generate highly detailed images, including portraits and scenes, and has received widespread recognition for its capabilities. On the other hand, Stable Diffusion made its model available as open-source, allowing anyone to access and use it.

MidJourney is another advanced text-to-image model that has garnered attention for its capabilities. It is considered by some to be the best of the three mentioned models. Don’t trust me on that, just admire these results.

MidJourney Landing Page UI design (source)
MidJourney Generated Image (source)
MidJourney Generated Image (source)

With the right prompt you can get some amazing results!

Unfortunately, its exclusive availability through its Discord channel has been a barrier to a wider adoption.

What is MidJourney?

In a nutshell, MidJourney is a text-to-image model, and app, that can generate amazing images based on a text description. Everything happens on Discord (I’ll show you how), and is accessible for a monthly subscription after a trial period.

It is owned by an organisation also calle MidJourney. In their own words: “Midjourney is an independent research lab exploring new mediums of thought and expanding the imaginative powers of the human species. We are a small self-funded team focused on design, human infrastructure, and AI. We have 11 full-time staff and an incredible set of advisors.”

Starting with MidJourney, Step by Step

It takes less than a minute, and 4 steps, to get you started!

(1) The journey starts here: https://www.midjourney.com/home/.

Click on “Join the Beta”

MidJourney’s Home Page

(2) Accept Midjourney’s invitation to Discord

Join MidJourney’s Discord

(3) Create a Discord account if you don’t have one yet

(4) Done! You are in on Discord! Congratulations!

MidJourney’s Discord view on Dec 20th 2022

Generate your First Image

We’re in, now what?

  • “Getting started instruction are available in their public documentation: https://midjourney.gitbook.io/docs/
  • Select a “newbies” channel, write “/imagine”, select the /imagine command, and then your prompt, e.g.,
    /imagine: the beautiful Black gods princess with Afro hair , fiery eyes, afro hair in shape of flowers

(Note that the first time discord will ask you to sign their ToS, and you’ll need to try again with /imagine+prompt)

Once your prompt is validated, you’ll have to wait a bit. The server will queue your prompt and execute it when compute resources become available. It will generate 4 variations of your prompt (2x2 grid). It usually take a minute or so.

Update & Save my images, Where to find my Gallery?

Variations & Up-scaling

As mentioned before, generating 4 variations of your image can take a minute or so.

Once ready, you’ll be able to either

  • (1) request variation of one of the generated image (V1, V2, V3, V4),
  • (2) or Upscale the one you like (U1, U2, U3, U4)

Let’s look at what we created with our prompt.

Prompt Inspiration: (source) ; Used Prompt: the beautiful Black gods princess with Afro hair , fiery eyes, afro hair in shape of flowers.

Upscale and Save in your gallery

We’ll go with U3. And here’s the result after a minute!

Upscaled and enhanced image on MidJourney!

Your Gallery is here: https://www.midjourney.com/app/

Save Images: You don’t have to do anything to save the results of your prompts to the gallery, it’s automatic and you’ll find there both the grid (2x2) results as well as the variations and upscaled versions.

A few more notes worth mentioning

  • Pricing: When you sign-up you get some free compute equivalent to 25 images (prompt, variation, upscale). If you want to keep using MidJourney, you’ll have to subscribe. It starts at $8/month (if billed annually). See Table below.
  • LowCode integration: Interacting with MidJourney means integrating with Discord. The good news is that Discord is widely integrated in LowCode automation tools. But we’ll get to that in the third part of this series ;)

Have fun with MidJourney! In the second part, we’ll right the book on Prompting. Everything you need to know to get the most of this tech ;)!

If you liked this post, and want to see more, please consider supporting us: 🔔 clap & follow 🔔

--

--

BoredGeekSociety

Wassim Jouini, CTO and Head of AI @LegalPlace. 12yrs+ Building AI & Automation products for Scale-Up Startups | Ph.D. AI