How to Use MidJourney for Beginners: Step-by-Step Tutorial
Complete beginner's guide to MidJourney. Learn how to create stunning AI art from signing up to writing your first prompts, with real examples.
How to Use MidJourney for Beginners: Step-by-Step Tutorial
Let me be real with you. When I first heard about MidJourney, I thought it was one of those overhyped tech things that would fade in a few months. I was wrong. Very wrong. MidJourney has become one of the most powerful creative tools available to anyone with an internet connection — and you do not need any design experience to use it.
If you have ever looked at those jaw-dropping AI-generated images online and thought "I have no idea how people make those," this guide is for you. By the time you finish reading, you will have created your first AI image and understand exactly how to make more.
No technical background needed. No art degree required. Just you, a Discord account, and about ten minutes to get started.
---
What Exactly Is MidJourney?
MidJourney is an AI image generation tool that turns text descriptions into images. You type a description of what you want — say, "a golden retriever wearing a space suit on Mars" — and MidJourney creates that image from scratch in about 60 seconds.
It is not pulling images from Google. It is not stealing from artists' portfolios. It is generating entirely new images based on patterns it learned during training on a massive dataset of images and their descriptions.
Think of it like this: MidJourney learned what "dog" looks like, what "space suit" looks like, what "Mars" looks like — and then it figured out how to combine those concepts into something that has never existed before.
The results can be absolutely stunning. We are talking about photorealistic images, concept art that looks like it took weeks to create, illustrations that could hang in a gallery, and abstract pieces that make you feel something. All from a text description.
Why MidJourney Stands Out
There are several AI image generators out there — DALL-E, Stable Diffusion, Firefly, and others. So why does MidJourney get so much attention?
- Image quality — MidJourney consistently produces some of the highest quality AI images available. The aesthetic quality is often described as "default beautiful."
- Artistic range — From photorealistic portraits to fantasy concept art to minimalist logos, MidJourney handles diverse styles remarkably well.
- Active community — The Discord community is massive, which means you can learn from millions of other users' prompts and results.
- Constant updates — The team ships improvements regularly. Each version brings noticeably better results.
The trade-off? MidJourney requires a paid subscription (no free tier anymore) and runs through Discord, which can feel unusual at first. But trust me — once you get the hang of it, the Discord workflow becomes second nature.
---
MidJourney Subscription Plans
Before you can generate any images, you need a subscription. Here is what is available:
Basic Plan — $10/month
- About 200 image generations per month
- Access to the member gallery
- General commercial usage rights
- Perfect for casual users and beginners who want to experiment
Standard Plan — $30/month
- 15 hours of Fast generation time
- Unlimited Relaxed generations (slower but no limit)
- All Basic features
- Best value for regular users who generate images frequently
Pro Plan — $60/month
- 30 hours of Fast generation time
- Unlimited Relaxed generations
- Stealth mode (your images stay private)
- Great for professionals and content creators
Mega Plan — $120/month
- 60 hours of Fast generation time
- Unlimited Relaxed generations
- Stealth mode
- For heavy users and businesses
My recommendation for beginners? Start with the Basic Plan at $10/month. It gives you enough generations to learn the ropes and decide if MidJourney is something you want to invest more time in. You can always upgrade later.
---
Getting Set Up: Your First 5 Minutes
Here is the step-by-step process to go from zero to your first AI-generated image.
Step 1: Create a Discord Account
MidJourney operates through Discord, a messaging platform. If you do not have Discord yet:
- Go to [discord.com](https://discord.com)
- Create a free account
- Download the app (desktop or mobile) or use the web version
If you already have Discord, you are ahead of the game.
Step 2: Subscribe to MidJourney
- Visit [midjourney.com](https://midjourney.com)
- Click "Join the Beta" or "Sign In"
- Connect your Discord account
- Choose your subscription plan and complete payment
Step 3: Access the MidJourney Bot
Once subscribed, you have two options for generating images:
Option A: MidJourney Discord Server
- Join the official MidJourney Discord server
- Navigate to any of the `#newbies` channels
- This is where you will type your first command
Option B: Direct Message the Bot
- Find the MidJourney Bot in the server member list
- Send it a direct message
- This gives you a private space to generate (recommended for beginners)
Option C: MidJourney Website
- You can also generate images directly at midjourney.com
- The web interface is more intuitive for beginners
- This is becoming the preferred method for many users
Step 4: Generate Your First Image
Here is the moment you have been waiting for. In Discord, type:
`/imagine prompt: a cozy coffee shop on a rainy evening, warm light glowing from the windows, watercolor style`
Press Enter. Wait about 60 seconds.
MidJourney will generate a grid of four different images based on your description. And just like that — you have created AI art.
---
Understanding the Image Grid
When MidJourney finishes generating, you will see a 2x2 grid of four image variations. Below the grid, you will notice two rows of buttons:
The U Buttons (U1, U2, U3, U4)
These stand for Upscale. Clicking U1 upscales the top-left image, U2 upscales the top-right, U3 upscales the bottom-left, and U4 upscales the bottom-right. Upscaling gives you a higher resolution, standalone version of that image.
The V Buttons (V1, V2, V3, V4)
These stand for Variations. Clicking V1 creates four new variations based on the top-left image. This is incredibly useful when you like the general direction of one image but want to explore similar options.
The Re-roll Button
The circular arrow button re-runs your entire prompt, generating four completely new images. Same description, different results — because there is always an element of randomness in AI generation.
Ready to master AI?
Our Complete AI Bootcamp covers prompt engineering, ChatGPT, MidJourney, vibe coding, AI agents and more — with 110+ video lessons and 2,000+ prompts.
Pro tip for beginners: Do not stress about getting the perfect image on your first try. The workflow is meant to be iterative. Generate, pick your favorite, create variations, refine. Most great MidJourney images come from 3-5 rounds of iteration.
---
The Key Parameters That Change Everything
Parameters are commands you add to the end of your prompt to control specific aspects of the generation. They start with `--` and they are what separate basic users from people who consistently get amazing results.
--ar (Aspect Ratio)
By default, MidJourney generates square images (1:1). But you probably want different dimensions:
- `--ar 16:9` — Widescreen, perfect for desktop wallpapers and YouTube thumbnails
- `--ar 9:16` — Vertical, ideal for phone wallpapers and Instagram stories
- `--ar 4:3` — Classic photo ratio
- `--ar 3:2` — Standard photography ratio
- `--ar 2:3` — Portrait orientation
Example: `/imagine prompt: a mountain landscape at sunset with purple clouds --ar 16:9`
--v (Version)
This selects which version of MidJourney's AI model to use. Higher versions generally produce better results:
- `--v 6.1` — The latest version with the best quality and prompt understanding
- `--v 6` — Previous version, still excellent
- `--v 5.2` — Older but some users prefer its aesthetic for certain styles
My advice: Stick with the default (latest version) unless you have a specific reason to use an older one.
--style
Controls the aesthetic approach:
- `--style raw` — Less "beautified," more literal interpretation of your prompt. Great when MidJourney's default aesthetic is too polished for what you need.
--stylize (or --s)
Controls how much MidJourney applies its own artistic interpretation. The value ranges from 0 to 1000:
- `--s 0` — Very literal, follows your prompt closely
- `--s 100` — Default balance
- `--s 250` — More artistic interpretation
- `--s 750` — Heavily stylized, MidJourney takes more creative liberty
Example: `/imagine prompt: a red barn in a wheat field --s 50` vs `--s 750` will give you dramatically different results. Try both!
--chaos (or --c)
Controls how varied the four images in your grid will be:
- `--c 0` — All four images will be quite similar
- `--c 50` — Moderate variety
- `--c 100` — Maximum variety, each image could look very different
This is useful during exploration. When you are still figuring out what you want, use higher chaos. When you are refining toward a specific look, use lower chaos.
--no (Negative Prompting)
Tells MidJourney what to exclude from the image:
- `--no text, words, letters` — Removes text from the image (very useful)
- `--no people` — Removes humans from the scene
- `--no watermark` — Helps avoid watermark-like artifacts
---
10 Beginner-Friendly Prompt Examples
Here are ten prompts you can copy and paste right now to start experimenting. Each one demonstrates a different style and use case:
1. Cozy Interior Photography
`/imagine prompt: a Scandinavian living room with warm sunlight streaming through large windows, minimalist furniture, plants on the windowsill, soft neutral tones, interior design photography --ar 16:9`
2. Epic Landscape
`/imagine prompt: dramatic mountain peaks above the clouds at golden hour, snow-capped summits, cinematic lighting, aerial photography --ar 16:9 --s 200`
3. Character Illustration
`/imagine prompt: a friendly robot librarian organizing books on tall shelves, cozy library atmosphere, warm lighting, Pixar style illustration --ar 3:4`
4. Product Photography Style
`/imagine prompt: a ceramic coffee mug on a wooden table, steam rising, morning light, blurred garden background, product photography, shallow depth of field --ar 4:5`
5. Fantasy Art
`/imagine prompt: an ancient tree with glowing golden leaves in an enchanted forest, magical particles floating in the air, soft ethereal light, fantasy concept art --ar 2:3`
6. Abstract Art
`/imagine prompt: flowing liquid marble in deep blue and gold, abstract art, organic shapes, luxurious texture, high detail --ar 1:1 --s 500`
7. Retro and Vintage
`/imagine prompt: a 1960s diner at night, neon signs, classic cars parked outside, nostalgic atmosphere, retro Americana, cinematic film photography --ar 16:9`
8. Minimalist Logo Concept
`/imagine prompt: minimalist logo design for a mountain coffee company, clean lines, simple geometric shapes, flat design, white background --no text, words, letters`
9. Watercolor Illustration
`/imagine prompt: a cat sitting in a window watching rain fall outside, loose watercolor painting style, soft muted colors, gentle brushstrokes, artistic --ar 4:5`
10. Photorealistic Portrait
`/imagine prompt: portrait of an elderly craftsman in his workshop, weathered hands, warm directional lighting, shallow depth of field, 85mm lens, documentary photography --ar 2:3 --s 150`
Try each one. Then change a few words and see how the results shift. That experimentation is how you develop your instinct for prompting.
---
Organizing Your MidJourney Images
Once you start generating, you will quickly accumulate a lot of images. Here is how to stay organized:
On the MidJourney Website
- Visit midjourney.com and log in
- All your generated images are saved in your gallery
- You can search through your past generations
- Use the "Liked" feature to bookmark your favorites
Downloading Your Images
- Click on any image in your gallery or in Discord
- Select the highest resolution version
- Save it to organized folders on your computer
Recommended Folder Structure
Create a system that works for you. Something like:
- MidJourney (main folder)
- Experiments (for exploration)
- Favorites (your best work)
- Prompts (text file with prompts that worked well)
The prompts file is crucial. When you get a great result, save the exact prompt that created it. Future-you will thank present-you for this habit.
---
Common Beginner Questions
"Why do my images look different from what I imagined?"
This is normal. AI image generation involves randomness, and MidJourney interprets prompts based on its training data. Two things help: be more specific in your descriptions, and use the parameters we discussed. Also, iteration is key — use the V buttons to refine images that are close to what you want.
"Can I use MidJourney images commercially?"
Yes. All paid subscription plans include general commercial usage rights. You can use your generated images for social media, blog posts, marketing materials, products, and more. However, you should review MidJourney's current Terms of Service for any specific restrictions that may apply to your use case.
"Why does MidJourney sometimes add weird text to images?"
AI image generators are not great at rendering text. If you see garbled letters appearing in your images, add `--no text, words, letters, writing` to your prompt. This usually helps.
"How do I get consistent characters across multiple images?"
This is one of the trickier aspects of MidJourney. Some approaches:
- Use very detailed character descriptions in every prompt
- Use the `--cref` (character reference) feature to maintain consistency
- Use the `--sref` (style reference) feature to maintain visual consistency
- Use seed numbers (`--seed`) to reduce randomness
"Is it better to write long or short prompts?"
Both can work, but for beginners, medium-length prompts tend to perform best — about 15-40 words. Too short and MidJourney fills in the gaps with its own interpretation. Too long and the AI might ignore some of your details. Focus on the most important elements: subject, style, lighting, mood.
"Can I edit or modify a generated image?"
Yes. MidJourney offers several editing features:
- Vary Region — Select a specific area of an image and regenerate just that portion
- Zoom Out — Expand the canvas to show more of the scene
- Pan — Extend the image in a specific direction
These tools give you more control over the final result without starting from scratch.
---
Your Action Plan: What to Do Right Now
Here is exactly what I recommend doing in your first session with MidJourney:
1. Sign up and subscribe to the Basic Plan ($10/month)
2. Run 3-5 of the example prompts I gave you above, exactly as written
3. Pick your favorite image from those results and create variations using the V button
4. Experiment with one parameter — I recommend starting with `--ar` since it has the most immediately visible effect
5. Save your best prompt in a text document for future reference
6. Try writing your own prompt from scratch — describe something you actually want to see
Do not overthink it. The best way to learn MidJourney is to generate a lot of images and observe what works. You will develop an intuition faster than you think.
---
What Comes Next?
You now know more than 90% of people who have "tried" MidJourney. Most people type a few random words, get mediocre results, and give up. You understand the grid system, parameters, prompt structure, and the iterative workflow that leads to great images.
But this is just the beginning. There is so much more to explore:
- Style references that let you match a specific aesthetic
- Image prompts that let you use an existing image as inspiration
- Character references for consistent characters across generations
- Advanced prompting techniques like multi-prompts and prompt weighting
If you found this guide helpful, check out our other articles on writing professional MidJourney prompts and making money with AI art. And if you are serious about mastering AI tools, our complete bootcamp covers everything from prompt engineering to building real projects with AI.
The creative tools are in your hands now. Go make something you are proud of.
Written by Saad A
AI Expert Instructor with experience at Deloitte, PwC, BMO, and Microsoft. Teaching 24,318+ students worldwide.
Ready to master AI?
Our Complete AI Bootcamp covers prompt engineering, ChatGPT, MidJourney, vibe coding, AI agents and more — with 110+ video lessons and 2,000+ prompts.