Midjourney Review 2023 – A New Era of AI-Driven Marketing

The world of digital marketing is constantly changing and evolving, so staying at the forefront of innovation is crucial. Enter Midjourney, a trailblazer in the integration of artificial intelligence (AI) with the creative process. This generative AI program, developed by the San Francisco-based research lab Midjourney Inc., is taking the advertising industry by storm, promising to transform the way marketers brainstorm ideas and create original content.

Founded by David Holz, the co-founder of Leap Motion, Midjourney uses sophisticated AI algorithms to generate images from natural language descriptions, a capability that has significant implications for marketing professionals. This Midjourney review aims to explore the platform’s potential as an innovative marketing tool, the impact it’s already had on the advertising landscape, and what the future might hold for this remarkable piece of technology.

Midjourney is more than just an AI program, it’s a vision of a new era in AI-driven marketing. As it continues to evolve and release new model versions, Midjourney is not only redefining the boundaries of what AI can do, but also reshaping the ways marketers engage with audiences and create compelling advertising content.

Stay with us as we delve into the fascinating journey of Midjourney, a true testament to the limitless potential of AI in the realm of marketing.

What is Midjourney

Midjourney is a Generative AI program and service that generates images similar to Stable Diffusion and OpenAi’s DALL-E. This text-to-image AI tool entered open beta on July 12, 2022, and users can create artwork with Midjourney using Discord bot commands. Users generate images by using the “/imagine” prompt and typing in what they want in simple terms extended with some special phrase; then, the bot returns a set of four images, from which users can choose which images they want to upscale or which images they want more versions.

It is also possible to generate images corresponding to the style of famous painters, graphic artists, and photographers, and thus, for example, you can have a brand new Van Gogh-style painting on a picture. Midjourney has faced criticism from some artists for allegedly reducing the value of their original creative work by utilizing it in the training set. However, Midjourney’s terms of service address this concern by including a DMCA takedown policy. Artists can utilize this policy to request the removal of their work from the set if they suspect copyright infringement.

Midjourney Pricing

In addition to the free trial version, you can choose out of three paid subscriptions, depending on the future usability of the images and the amount of hardware resources (GPU time) guaranteed to you. The first Midjourney plan where you can use the created image for commercial purposes is the Basic Plan which costs $10 per month or $96 per year. To start with a free version, you just need to create a free Discord account, log in to any servers, and you can start promoting and generating pictures.

• Free Trial – 0,4 hr/lifetime Fast GPU Time – You must give appropriate credit in case of publishing created pictures and not using them for commercial purposes. Usage right: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) – you need to consider that quite often, your prompt is rejected due to high server usage.
• Basic Plan – 3,3 hr/month Fast GPU Time – Commercial usage of created pictures for companies making under $1M USD yearly gross revenue. As this plan not has relaxed GPU Time – like the upcoming ones – this plan has a ~ 200 generation/month limitation.
• Standard Plan – 15 hr/month Fast GPU Time – Commercial usage of created pictures for companies making under $1M USD yearly gross revenue.
• Pro Plan – 30 hr/month Fast GPU Time – Commercial usage of created pictures for companies making over $1M USD yearly gross revenue.

How to Get Started and make your first Midjourney Login

As mentioned before, the system uses a Discord bot to get prompts for image creation and to show the generated pictures. If you don’t have an account, then you need to create one for free here. As soon as your Discord account is available, visit the Midjourney website and click on the Join the Beta button on the right-hand side of the screen.

Now you are invited to the Discord server, so you are able to log in and find a public room and start the text-to-image journey.

Discord User Interface

Discord is a digital communication platform that enables text, voice, and video interactions, primarily aimed at gaining communities, but also widely used for various other types of communities, just like in our case for Midjourney. If you want to jump right into generating your first image, all you have to do is look for a friendly newcomer room on the left-hand side of the Discord screen and click on it. In the picture below, I just chose the newbies-31 room, but any room marked with a different number will also be perfect to start with.

The essence of the concept is open communication and learning from the images, and from the commands that others used to generate those. For this reason, these chat rooms are open, and you can see each other’s interactions on the server, and you can even download pictures created by others.

To create the first images, you need to use the “/imagine” command in the Message field at the bottom of the screen. When you have entered this, a new field, named prompt will appear, where you can freely state what kind of picture you want to have made. You can find more complicated expressions and attributes in a later chapter of this article, now, it will be enough just to enter a simple sentence. Let’s see an example:

About a minute after issuing the command, 4 different but still similar images are created based on the described instructions. By clicking on the images, you can view them larger, and if you like one of them and want to continue using it, save it; then you have two options.

• U1, U2, U3, U4 – Upscale the selected image – by default, the aspect ratio is 1:1, and the upscale will be 1024 px * 1024 px (if you ask the pictures in 16:9, then the upscale will be 1456 px * 816 px.
• V1, V2, V3, V4 – Variations of the selected image – in this case, you will get 4 other variants of the selected picture where only minor changes will be made.

numbering starts from the top left, then the top right, bottom left, and bottom right.

• Or, with the refresh button, you can as a completely new set of 4 images with the same instruction. Click here to see the pictures generated by that sentence.

To download any of the newly created images, upscaled or not, you should click on the picture to get the full size and right-click Save Image; however, all images are immediately available on the Web App.

Midjourney Web App

To open the web app, use the https://www.midjourney.com/app/ URL, and log in with your Midjourney account.

In addition to the obvious functions, like account and subscription management, the most important reason to use this interface is that you can organize and download the files from here. You can filter only for those images that you have upscaled, but here you can also view those that are only in the “offer phase” and 4 versions are on the same picture.

Midjourney prompts

Basically, a complex image creation prompt can be divided into 3 parts, I will start describing in the order that they must follow each other in the command.

Image Prompts

Incorporating images into a prompt can greatly impact a job’s overall appearance, including its style, composition, and color scheme. You have the option to use image prompts on your own or alongside text prompts. Experiment with various image styles to achieve the most compelling results. To add images to your prompt, simply enter or paste the web address where the image is stored online. Ensure that the address ends with a file extension such as .png, .gif, or .jpg. Once you’ve added the image addresses, include any supplementary text or parameters to finalize the prompt.

Midjourney offers a solution to create their online image direct link, if you use Discord, and copy past your file to the chat. When it’s got uploaded, you just need to use right-click on the image and “Copy Link” or “Copy Image Address”, depending on whether you use the Desktop app or Web App.

Text Prompts

This is a simple text description of what you want to see in the picture. As there is a pre-programmed Midjourney basic style, a single word or even an emoji is enough to generate an image. Of course, a more detailed description will result in more unique images, although the more details we describe, the less prominent each comment will be. It’s not always easy to find the balance, but can easily get experience to be more comfortable with prompting.

If we have not defined some aspect of the image, then the AI will define it according to the basic settings, and best fit to our composition. For this reason, if we leave out an important detail, we may not get the picture that we expected. I suggest considering the following aspects, and, if possible, adding them to the text prompt.

• Subject: person, animal, object, landscape, etc.
• Medium: photo, oil painting, doodle, graphics, etc.
• Environment: indoors, forest, city, sea, moon, Roxford, etc.
• Lighting: ambient, neon, studio, soft, etc.
• Color: pastel colors, autumn colors, shades of green, bright, etc (pay attention to word order, AI tends to focus on the color written earlier in the prompt).
• Mood: calm, fearful, fun, energetic, nostalgic, romantic, etc.
• Composition: birds-eye, portrait, closeup, etc.

Parameters

The parameters are always at the end of the command, and all start with a signal character “–“.

• Aspect Ratios – example: “–ar 16:9” for docs and articles or “–ar 1:1” for Insta story.
• No – example: “–no green color” or “–no cars.”
• Stop – example: “–stop 90” will finish on 90%, and the image will be blurrier, and less detailed.
• Chaos – example “— chaos 10” higher value results unusual generations 10% unexpected factor.

Conclusion

All in all, it’s extremely detailed with plenty of custom settings, providing unlimited image creation options from the hyper-realistic to a fantasy picture. We can get perfect pictures in the case of almost all subjects, although there are a few areas where we found room to improve, such as the illustration of hands and fingers, or handling of numbers and letters.

The operating interfaces are clear and easy to understand, and the most important commands and settings are easy to learn. The free account option is very friendly, but the user in this package almost never gets server time, especially in the frequented hours. If you need an image quickly, or if you want inspiration for an illustration, even for equipment or an object, you can rely on Midjourney’s help; you won’t be disappointed.

For as little as $8 a month, you can get an almost all-knowing text-to-image solution for creating both realistic and futuristic images. There are unlimited opportunities to creat unique images and illustrations, which can be used not only for direct use, but also for inspiration.