Rather than flipping pages on Pinterest for inspiration, what if you can give an AI bot what you’re looking for and have it provide you with something no one has seen before? This is the power of MidJourney’s AI/ML platform! This popular new tool is all the rage right now and this blog post takes you through MidJourney workflow, like your first steps with it and what it can do.
What is MidJourney?
MidJourney is a lab that allows users to generate images using AI, artificial intelligence. You may have seen some people posting a four-tiled picture and sharing some keywords describing the depiction. Well, most likely, that was AI through MidJourney. Users can give the Bot a prompt, and it will generate visuals from it. So what’s the deal? What makes this so unique? Well, these images are entirely original and generated within a minute. It’s accessible to the general public and can save time and effort in creating fresh content.
What is MidJourney used for?
There is an endless possible use for MidJourney: It is great for people trying to find inspiration, wanting copyright-free content, creating references for client decks, creating texture materials for 3D assets, and much more.
The Service
There are different AI labs available, yet MidJourney has been one of the more popular applications. It has a free trial version and additional tier subscription-based options.
In the plans below, each time amount is described in minutes and it takes approximately ~1 minute per image set. So for example, 25 minutes would be mean about 25 jobs / image sets.
MidJourney Free Trial:
- price: free
- fast GPU time: 25 minutes / lifetime
- personal bot chat: no
- private visibility: no
Basic:
- price: $10/ month
- fast GPU time: 200 minutes / month
- personal bot chat: yes
- private visibility: for $20/ month
- can give five friend invites
Standard:
- price: $30/ month
- fast GPU time: 15 hours / month
- personal bot chat: yes
- private visibility: for $20/ month
- can give five friend invites
Corporate:
- price: $600/ month
- fast GPU time: 120 hours / year (so this would mean ~7,200 image sets)
- personal bot chat: yes
- private visibility: yes
How to join?
There are two ways to experience the tool. In the Discord server and web app. You can find a gallery of your work and other users’ creations on the web app.
As of July 2022, MidJourney Beta requires people to apply for the service or be invited by another subscribed user.
MidJourney Workflow
Starting with a Free trial
After downloading the Discord app or logging in through a browser, you can join the server linked below. Then you can select one of the #Newbies channels from the sidebar.
To start a prompt, type in /imagine then press spacebar, tab, or enter to create a prompt field like below. Once you have the prompt field, you can type in whatever you like and hit enter again!
Wait as the MidJourney bot processes the request. The user will then be shown a 4 image grid with a few options below that are UI, U2, U3, U4, V1, V2, V3, V4.
The U is for upscaling, and V is for variation.
When selecting the U option, you can upscale the image to and if you want a variation of it or ‘Upscale to Max’. This process takes whichever of the 4 images you had above and provides you a much more detailed version of it (by upscaling it).
When selecting the V option, you’ll get variations based on that selection. The numbering corresponds to top-left (1), top-right (2), bottom-left (3), and bottom-right (4). Below is an example of upscaling one of the options above:
Get Our 7 Core TouchDesigner Templates, FREE
We’re making our 7 core project file templates available – for free.
These templates shed light into the most useful and sometimes obtuse features of TouchDesigner.
They’re designed to be immediately applicable for the complete TouchDesigner beginner, while also providing inspiration for the advanced user.
Saving the Images
You can click on the image to have it full size and then right-click and choose ‘Save image’ to save on your local device. The image will automatically be visible in your website gallery.
You can also send your generated images to yourself by reacting with the✉️ emoji on the image.
The automatic image resolution is 1024 by 1024 pixels.
Note : there are PG -13 rules of content generator. Some words are banned, such as ‘lust.’
Common Commands
/imagine: creates images from text. Provides grid of 4 images in ~50 seconds
/info : shows information about your profile and plan, and usage
/invite: generate a link to send to your DM to send to someone to join the server. (limit to 5 invites per subscribed user)
/ideas : give some random ideas for prompt
/help : display bot options for handy references
/subscribe : get a link to the subscription page
/fast and /relax: toggle between fast and relax mode. fast mode will be incrementally billed, whereas relax mode does not cost but takes a long time to generate.
/private and /public: toggle between ‘private’ and ‘public’ mode if you want to hide your show your image in the gallery. (costs an extra $20/ month for private option)
Private mode hides images and prompts from the gallery. This might be useful if you’re doing a lot of client work and want to keep your generated content behind closed doors.
Prompt Parameters
--w manually set the pixel width of images. (work better as a multiple of 64)
--h manually set the pixel height of images ( work better as multiple of 64)
--seed sets a random seed
--no exclude specific keywords from the prompt results, for example --no trees will remove any trees from the results
--video saves a progress video
--hd creates larger images, best for abstraction and landscape prompt
--aspect set a desired aspect ratio for exapmle --aspect 16:9
You can type these parameters after your prompt and separate them with a comma. For example here’s a prompt for meditative content with no humans:
/imagine prompt: meditation, peace, joy, water, --no humans
Resouces for Prompts:
What are the concerns of AI?
AI has been there for decades, yet the use of AI is expanding into a more diverse pool of industries. Questions of ethics, authorship, copyright, and ownership of creative content will still need to be explored. Having AI replacing jobs is a genuine concern. There are agencies replacing concept art teams with MidJourney. It will be a shift in the landscape of the creative industry. Where will the future hold? That we will have to wait to see.
MidJourney Workflow Wrap up
Whether you’re just starting your explorations of creative AI/ML platforms or you’re a seasoned veteran, MidJourney provides a lot of value in an easy to access package. Especially since there’s a free tier, I’d recommend everyone give it a try, if nothing else to know what is on the forefront of AI/ML-based creation tools. You might even end up using it more than a few times and incorporating it into your workflows or inspiration sessions!
