Your Simple Guide to Exploring the Midjourney Website
- kylixie
- Apr 30, 2025
- 10 min read
Getting started with the Midjourney website might seem like a lot at first. There are many options and buttons. It can feel confusing. But don't worry. This guide will walk you through how to use the Midjourney website step by step. It's simple and easy to understand. By the end, you'll know how to use the main features.
Many users find the website easier than Discord for finding and sorting images. The layout is simple and helps you keep your work organized. If you have access to the website, give it a try.
Getting Started on Midjourney.com
Once you log in, you'll see the homepage. It starts in the Explore tab. Here you can see images others have made. On the left side are different tabs:
Explore: See what others are creating.
Create: This is where you'll make your own images.
Archive: Find and sort your past images.
Rate Images: Help improve Midjourney by rating images.
Feedback: Share ideas with the team.
You can also switch between light and dark modes using the button in the bottom left corner. Pick whichever you like best.
Creating Your First Images
Go to the Create tab. It looks like a paintbrush. This is your main work area. You type your prompts here. You can also change settings for your pictures.
Images you make are shown in rows. The prompt and settings are on the right. The images are on the left. You can scroll down to see old images you generated.
To start, use the Imagine prompt box at the top. Type what you want to create. For example, "photo of Lake Tahoe on a clear summer day, there's a man on a canoe."
After you enter the prompt, you will see the images being made. The prompt is on the right, and the image grid appears on the left as it finishes.
Making Variations
When your images are done, you can make new versions by hovering over them. You'll see buttons: Vary Subtle and Vary Strong.
Vary Subtle: Makes new images that look mostly like the first one. Small details might change, but the main look and colors stay the same.
Vary Strong: Creates versions with bigger changes. The layout or larger details can be different.
Comparing the results, Vary Subtle keeps things similar. Vary Strong shows more noticeable differences, like the shape of mountains or trees in the background.
Customizing Your Generations with Settings
Let's look at the Settings menu. It gives you many ways to change your images. For example, you could prompt for "a 3D render of an isometric Asian Shrine. Created in Blender."
Find the Settings icon. It looks like three bars with circles. It's in the image tab when you're creating.
Image Size (Aspect Ratio)
You can change the shape of your images. Pick vertical (portrait), square (default), or horizontal (landscape). You can also drag a slider to get other shapes.
See how the settings appear under your prompt? Like "--ar 2:3". This is a shortcut. '--ar' means aspect ratio, and '2:3' sets the vertical shape. All your custom settings show up here so you know what was used.
Aesthetics Settings: Stylization, Weirdness, Variety
These settings under Aesthetics change the look and feel.
Stylization (--stylize): How much creative freedom Midjourney uses. Low stylization stays close to your prompt but is less artistic. High stylization adds more imaginative details.
Weirdness (--weird): Makes images unusual or unconventional. The results might not match your prompt exactly but can be very interesting.
Variety (--chaos): Controls how different the images in your grid are from each other. A high value means the four images will look quite different. (This used to be called Chaos)
You can see the difference when using these. A high stylize value might add many details like trees or water to an image, while a low value looks basic. High weirdness can produce many visual styles for the same prompt. High variety might make the images in a grid look very different from what you asked for.
You can add these as shortcuts directly into your prompt. Type "--stylize" or "--weird" followed by a number. For example, "coffee shop generated with anime style art --stylize 300 --weird 20". The settings will then show up below your prompt.
Models and Modes
Midjourney has different versions (models). You can change the model you are using. Version 6 is the newest default and makes high-quality images. Niji 6 is great for anime and illustration styles.
Generation Mode is also important for specific art styles.
Standard Mode: Midjourney adds its usual enhancements.
Raw Mode: More directly follows your prompt with fewer automatic changes. If you want a specific feel, like Studio Ghibli, Raw mode might keep the style closer to the original even if it\'s less bright than Standard.
You can also control speed. Fast mode uses GPU hours. Relax mode is slower but doesn't use GPU hours. Turbo mode is fastest but uses GPU hours quickly.
Hidden Parameters
There are other helpful parameters you can use:
Negative Prompting (--no): Say what you *don't* want in the image. If you have a landscape prompt but don't want mountains, add "--no mountains" at the end of your prompt. Midjourney will try to avoid them.
Tile (--tile): Creates images that can repeat to make a pattern. Add "--tile" to your prompt. This is great for fabrics or wallpapers. You can download the image and use online tools like pycheung.com/checker to see the repeating pattern and download it.
Permutations ({...}): Saves time by running multiple prompts at once. Use curly brackets {}. Inside the brackets, list words or phrases separated by commas. For example, "{blue, orange, green, red} car" will generate images of blue, orange, green, and red cars in one job. You can use this for words or even parameters like stylize values.
Managing many prompts and variations can become time-consuming. Consider using the Midjourney Automation Suite from TitanXT to streamline your workflow and easily manage sets of prompts and parameters.
Searching Your Images
On every page, there is a search bar at the top right. This helps you find images quickly. Search for keywords from your prompts, like "dolphin". You can also search for settings, like "--ar 9:16", to find all images made with that aspect ratio.
Below the prompt on the right of each image set, some icons pop up when you hover. The swirling arrow is Rerun; it runs the exact same prompt again. The 'T' is Use; it copies the prompt and settings to the main prompt box so you can change them easily. You can also hide images to clear your workspace (they are not deleted). You can download the full image grid or open it in Discord. You can also copy the prompt or the image link.
Changing and Enhancing Your Images
Once you have a base image, you can change it in many ways. Click on an image in the grid to open the features menu on the right.
Vary Subtle/Strong: We covered this - makes new versions of the image.
Upscale (2x): Increases the image resolution. A 1024x1024 image becomes 2048x2048.
Upscale Subtle: Increases resolution while keeping details very close to the original.
Upscale Creative: Increases resolution and sometimes makes minor changes to improve details, like faces that look unnatural in the original generated image. If the original image is already good, Upscale Subtle is often better to avoid changes.
After applying changes, red text bars next to your prompt show what features were used, like 'Upscale' or 'Variation'.
Remixing Images
Remix changes the image while keeping some of its original structure or subjects. Use Remix Subtle or Remix Strong.
Remix Subtle: Copies the prompt and the image into the prompt bar, adding a Remix tag. You can then add new words to change the environment or style while keeping the general layout. For example, take an image of hikers and remix with "in snow covered mountains wearing winter hiking gear" to keep the hikers but put them in a snowy scene.
Remix Strong: Creates more different results than Subtle, but the influence of the original image (like subject position) is still visible.
Changing Viewpoint: Pan and Zoom
These features let you expand your image. Pan extends the image in one direction (up, down, left, or right). It changes the aspect ratio and adds to the image without changing the center picture. Zoom pulls back the camera and expands the image in all directions, filling the new area based on the image and prompt.
Change AR is similar to Pan but more flexible. Use a slider to pick a specific aspect ratio. A box shows the new border. You can pick if the original image is at the Start, Center, or End of the new expanded picture. This is like custom zoom in Discord but with more control where the base image sits. You can use Pan and Zoom multiple times.
Inpainting (Vary Region)
Vary Region lets you edit specific areas of an image. Select a part of the image using a box or by drawing. Then, in the prompt box that appears, describe what you want in that selected area. Midjourney will try to change only that part. You can right-click to undo a selection mistake.
This is great for adding things Midjourney doesn't usually make well. For example, you might generate an image of Spider-Man. Then use Vary Region to select where the head is and change the prompt just for that area to "the Joker". This can help get a character like the Joker wearing a Spider-Man costume.
Keeping track of all the subtle and strong variations, upscales, remixes, pan, zoom, and vary region jobs can be complex. Speed up your creative flow and manage image transformations more efficiently by using a tool like the Midjourney Automation Suite from TitanXT.
Other Image Options
When you open an image, you'll see icons for 'Use':
Image: Use the current image as an image prompt (it appears as an attachment in the prompt bar).
Style: Use the style of the current image as a reference for new images. This helps keep a consistent style across different creations.
Prompt: Copy and paste the current prompt to the main prompt box.
Other icons:
Down arrow: Download the image.
Magnifying glass: Find similar images made by other users.
Star: Favorite the image to find it easily later.
You can also Right Click any image in the Create tab to see most of these options quickly (Pan/Zoom are not available here).
Using Images as References
You can upload your own images to use as starting points or guides for new generations. This helps create Consistent Characters and Consistent Styles.
Upload an image using the plus '+' icon in the prompt bar or drag and drop it.
Image Describing
Hover over an uploaded image and click the 'i' icon. Midjourney will suggest words or phrases describing what the image looks like. This is useful if you want to generate something similar but don't know the right words to use.
Blending Images
Drag and drop multiple images into the prompt bar. Midjourney can blend them together. If the original images had a portrait shape, add "--ar 2:3" to help the blended result also be portrait. The more images you blend, the less the final image might look like any single original one.
Image Weight
When using an image prompt, Image Weight (--iw) controls how much the new images should look like the reference image. For version 6, it's a number from 0 to 3, with 1 being the default. A lover number (like --iw 0.3) means the reference image has less influence. This helps when you use a pattern image but want the main subject clearer, not covered by the pattern.
Consistent Style
Upload a style reference image or images. After uploading, hover over them in the prompt bar and select the paperclip icon. This tells Midjourney to use the image for style. Then add your prompt, like "retro wall art of a spaceship --ar 2:3". Your new images will try to match the style of the references. You can use multiple style reference images, like a collection showing a 'Rustic Countryside Vibe', to influence the style of different house rooms.
Consistent Character
This feature helps create photos of the same person in different situations. Generate a base image of a character. Drag it into the prompt bar. Hover over it and select the person icon.
Now, add a new prompt describing the situation, expression, or camera angle. For example, "side profile shot from the side, happy expression, garden outside during golden hour".
It might take a few tries with different reference photos to get the character to look right consistently. Sometimes it can make faces look like a drawing instead of a photo.
Character Weight
Character Weight (--cw) is a value between 0 and 100. It changes how much Midjourney copies the exact look of the character from the reference image. The default is 100, which tries to match the face, hair, and clothing closely. A lower value, like "--cw 10", gives you more freedom to change the character's clothes or hairstyle in the new prompt. If you use the default 100 weight, Midjourney might not change the clothing when you ask it to.
Achieving perfect consistency across characters or styles requires careful management of references, weights, and many prompt variations. Automate the process of generating and testing character and style prompts with the Midjourney Automation Suite from TitanXT to save time and improve results.
Organizing Your Images in the Archive Tab
The Archive tab is great for keeping your images tidy. It helps you find specific pictures easily. First, choose how you want to see images. The default is Square thumbnails, but Full layout shows the whole image. You can also change the size of the pictures displayed.
Filters and Search
Use the Filters tab to show only certain types of images. You could filter to only see images you have Upscaled or those with a vertical shape. Combine filters with the Search bar to narrow down more. Search for "tree" along with your Upscaled, portrait filter.
Folders and Smart Folders
Folders help you group your work. Create a new folder, give it a name like "ancient ruins", and drag images into it from the viewing area.
Smart folders are even faster. Create a folder and check the "smart folder" box. Enter keywords or parameters, separated by commas (like "Polaroid, Instax, Disposable camera"). The folder will automatically add any images that have those keywords in their prompts. It might take a moment and a page refresh to update.
Finding Inspiration
Go back to the Explore page to see what others have made. See popular images and use any picture there as a reference. You can copy their prompt and settings to your prompt bar.
The search bar on the Explore page helps you find examples of different styles or subjects, like "watercolor wall decor". You can also visit other users' profiles.
You can help improve Midjourney by rating images. This helps them train their models. Getting a high rating rank can even earn you free GPU hours.
The Feedback page lets you suggest new features. You can also see ideas other users have shared and what features are most wanted (these might be added soon). Check here for the latest Midjourney updates too.
Common Questions
Can I use my own images for Pan, Zoom, or Inpainting? Not right now. The features work best with images made by Midjourney.
Can I use my own photos for Consistent Characters? You can try, but it usually works much better with characters generated by Midjourney.
Conclusion
The Midjourney website offers many powerful tools for creating and managing your AI art. From basic prompting and parameter settings to advanced image manipulation, references, and organization features, there's a lot to explore. Spend time experimenting with these options to see what you can create.
For users looking to scale their creation process, manage large libraries of images, and automate complex prompt variations and consistent character/style generations, exploring tools like the Midjourney Automation Suite from TitanXT can take your workflow to the next level.
Comments