- Midjourney requires a paid subscription before you can create any images—log in at midjourney.com and choose a plan.
- Type your prompt in the Imagine bar at the top of the Create page, press Enter, and Midjourney generates four images.
- Refine results with upscalers, variations, zoom, pan, and the Editor; use Image Prompts and Style References to guide the AI with your own pictures.
- All your creations are saved automatically on the Organize page for easy browsing.
Midjourney is one of the most popular AI image generators, known for its ability to create detailed, artistic images from text descriptions. Unlike some AI tools that let you try before you buy, Midjourney requires a subscription before you can create anything. Once you’re in, the entire process happens on the web—no software to install and no command-line knowledge needed. This guide walks you through subscribing, writing your first prompt, refining the results, and organizing your work.
What is Midjourney?
Midjourney is an AI image generator that turns text prompts into pictures. You describe what you want—a sunset over mountains, a robot reading a book, a vintage poster for a coffee shop—and the AI creates four variations for you to choose from. It’s designed for creative work: concept art, illustrations, marketing visuals, or just experimenting with ideas.
The service runs entirely in your web browser at midjourney.com. You don’t download anything, and you don’t need a powerful computer—the image generation happens on Midjourney’s servers. Every image you create is saved to your account automatically.
For a detailed look at features, pricing, and how Midjourney compares to other tools, see our Midjourney review.
Step 1: Subscribe to Midjourney
Midjourney does not offer a free trial or a free tier. To create any images, you must subscribe first. Here’s how:
- Go to midjourney.com and click Sign In.
- Log in with your Google account or email.
- Once logged in, you’ll see the subscription plans. Select the one that fits your needs.
- Complete the payment, and your account is activated immediately.
Midjourney offers monthly and annual billing. Annual plans cost less per month. Check the official getting started guide for current plan information, as these change from time to time.
After subscribing, you land on the Create page, ready to generate your first image.
Step 2: Create Your First Image
The Create page is where you’ll spend most of your time. At the top of the page is the Imagine bar—a text field where you type your prompt.
A prompt is a short description of what you want to see. For example:
- “a quiet library filled with plants and sunlight”
- “a vintage travel poster for Tokyo”
- “a golden retriever wearing a wizard hat”
Type your prompt into the Imagine bar and press Enter. Midjourney takes about 30 seconds to a minute to generate a set of four images. Each one is a different interpretation of your prompt. You’ll see a progress bar while the AI works, then all four images appear together in a grid.
If you’re new to writing prompts, start simple. Describe the subject, the mood, and maybe a style (“watercolor painting,” “black and white photo,” “3D render”). For more detailed guidance on prompt writing and comparing AI image tools, see our beginner’s guide to making AI images.
Step 3: Refine Your Images
Once you have your four images, you can refine them using the tools that appear below the grid. Here’s what each tool does:
Upscalers
Click the Upscale button under any image to generate a larger, higher-resolution version. This is useful when you’ve found an image you like and want to save it at full quality for printing or publishing.
Variations
The Vary button creates four new images based on the one you selected. The AI keeps the overall composition and style but changes details. If you like an image but want to explore alternatives, variations are the fastest way to do it.
Zoom Out and Pan
Zoom Out expands the frame around your image, revealing more of the scene beyond the original borders. Pan shifts the frame in a specific direction—up, down, left, or right—without changing the content of the original image. Both tools are useful when the AI crops something you wanted to see.
Editor
The Editor lets you select a region of the image and re-generate just that part with a new prompt. For example, if you have a landscape but want to change the sky, you can paint over the sky area and tell the AI what to put there instead. This is one of the most powerful tools for fine-tuning details without starting over.
Personalization: Teaching Midjourney Your Style
Midjourney includes personalization features that learn your preferences over time. You can create a personalization profile by rating images—thumbs up for styles you like, thumbs down for styles you don’t. The AI uses these ratings to adjust future generations to match your taste.
You can also build moodboards by collecting images you like (your own or others’) into a board. Midjourney analyzes the colors, composition, and style of the board and applies that aesthetic to new prompts when you reference the board. This is useful if you’re working on a project with a consistent visual identity and want every image to feel like part of the same set.
Settings and Customization
Click the settings icon next to the Imagine bar to open the settings panel. Here you can adjust:
- Aspect ratio: Square, portrait, landscape, or custom dimensions.
- Model version: Midjourney releases new model versions over time. The default is the latest version, but you can switch to older versions if you prefer their style.
- Stylize: Controls how much the AI interprets your prompt artistically. Higher values produce more stylized, dramatic images; lower values stay closer to a literal interpretation.
- Chaos: Increases variety between the four generated images. Useful when you want very different options to choose from.
- Quality: Higher quality takes longer but produces more detailed images.
These settings apply to every new prompt until you change them again. Experiment with different combinations to see what works for your project.
Using Images as Input
You can guide Midjourney with images as well as text. This is especially useful when you have a reference photo, a color palette, or a specific style you want to match.
Image Prompts
Upload an image alongside your text prompt. The AI uses the image as a visual starting point, combining its content with your written description. For example, you could upload a photo of a building and write “at sunset, painted in watercolor” to transform the photo into a painting.
Style References
Style References focus on the aesthetic of an image—the colors, lighting, and artistic treatment—without copying the content. Upload a reference image and write a new subject, and the AI applies the reference’s style to your new prompt. This is powerful for maintaining a consistent look across many images.
Starting Frame (Video Generation)
Midjourney has added the ability to use an image as the first frame of a short video. Upload an image, and the AI animates it. Check Midjourney’s official documentation for current details on this feature, as video capabilities are evolving.
Edit Model
The Edit model is similar to the Editor tool but works with uploaded images. You can bring in your own photo, select a region, and ask the AI to change just that part. It’s useful for touching up photos or blending AI elements into real images.
Organizing Your Creations
Everything you generate is saved automatically. Click the Organize tab at the top of the page to see all your images in a gallery. You can:
- Filter by date, rating, or folder.
- Create folders to sort images by project.
- Search by the prompt you used.
- Download images individually or in bulk.
Your images are private by default, but you can choose to make them public if you want to share them in Midjourney’s community gallery.
How Midjourney Compares to Other AI Image Generators
Midjourney is one of several AI image tools available in 2026. It’s known for highly detailed, artistic results and strong stylization, making it popular with illustrators and designers. However, it requires a subscription from the start, unlike some competitors that offer free tiers.
For a broader comparison of AI image tools, including free options, see the best AI image generators right now and free AI image generators.
If you’re considering other AI assistants that also handle images alongside text, the best ChatGPT alternatives covers a range of options for different use cases.
Frequently Asked Questions
Can I use Midjourney for free?
No. Midjourney does not offer a free trial or free tier. You must subscribe to a paid plan before you can generate any images. Visit the official pricing page at midjourney.com for current plan options and costs.
Do I need to install any software to use Midjourney?
No. Midjourney runs entirely in your web browser at midjourney.com. You don’t download anything, and the image generation happens on Midjourney’s servers. All you need is an internet connection and a modern browser.
Can I use Midjourney images commercially?
It depends on your subscription plan. Some plans grant full commercial rights to the images you create; others do not. Read the terms of service on your account page or check Midjourney’s documentation for details on what your plan allows.
What does the Stylize setting do?
Stylize controls how much artistic interpretation the AI applies to your prompt. A low stylize value produces images that stick closely to your description. A high stylize value gives the AI more freedom to add dramatic lighting, composition, and artistic flair. Adjust it in the settings panel to match your project’s needs.
How do I write better prompts?
Start with a clear subject, add details about mood or lighting, and include a style reference if you have one in mind (“oil painting,” “minimalist line art,” “cinematic photo”). Avoid overly long prompts—Midjourney works best with focused descriptions. For more tips, see how to make AI images: beginner’s guide.
Can I edit images after they’re generated?
Yes. Use the Editor tool to select a region of an image and re-generate just that part with a new prompt. You can also use the Edit model to bring in your own photos and blend AI changes into them. Both tools let you refine details without starting over from scratch.