Midjourney User Guide: Join Discord, Generate Your First Image, and Common Commands

This Midjourney tutorial is designed to address the three things beginners most often get stuck on: how to get into Discord and find a usable channel, how to use /imagine to generate your first image, and how to continue upscaling and editing after the image is generated. Follow the steps below and you can usually get Midjourney up and running—and reuse it reliably—within about ten minutes.

1. Preparation Before You Start: A Discord Account and Midjourney Login

Midjourney mainly interacts through Discord, so first register and log in to Discord, then go to the Midjourney website and click “Join the Discord” to join the official server. After entering the server, you usually need to agree to the rules or complete a simple verification. Once finished, you’ll be able to use Midjourney commands normally in the channels.

It’s also recommended that you log in to the Midjourney website and authorize linking with the same Discord account, which makes it easier to view your personal gallery and task history later. After linking, the Midjourney images you generate in Discord will generally also appear in your web gallery for easier downloading and management.

2. Your First Image: Use /imagine in a Newbies Channel

After entering the official Midjourney server, search for or find a beginner channel with “newbies” in the name in the sidebar. In the input box, type “/imagine”, select the command that pops up, and then write your description in the prompt to start a task—this is the most commonly used entry point for Midjourney.

You can write prompts directly in Chinese, but for more consistent results, try to clearly specify the subject, style, camera, and lighting—for example: “indoor portrait, film texture, side backlight, shallow depth of field.” If you want a fixed aspect ratio, add a parameter at the end such as “--ar 3:4”, and Midjourney will compose according to that ratio.

3. How to Modify After Generating: U/V, Rerolls, and Extending the Frame

After Midjourney generates a 2×2 grid, you’ll see U and V buttons: U is usually for upscaling a specific image, and V is for creating similar variations based on a specific image. If you’re unsure about direction, a more reliable workflow is V first, then U: lock in the style first, then upscale the one you like best.

After upscaling, you’ll often see fine-tuning options like “Vary (Strong/Subtle)”, which adjust details while keeping the subject. When you need to expand the scene, choose “Zoom” or continue describing “extend the scene outward.” All of these actions belong to the same Midjourney task chain, making it easy to iterate step by step until you reach the final image.

4. Common Commands and Quick Troubleshooting: /settings and Queue Messages

If you want to quickly check your current mode and common toggles, use “/settings”, where you can switch certain generation preferences (the specific options may change as Midjourney updates). If you’re temporarily out of ideas, you can use “/describe” to upload an image and have Midjourney reverse-engineer prompt suggestions, then edit the descriptions you like into your own version.

If the queue is long or responses are slow, first confirm that the channel where you sent the command allows Midjourney, and that your network connection to Discord is stable. If it says the command is invalid, it’s usually because you didn’t select the slash command correctly or you haven’t completed server verification; re-enter the channel and type “/imagine” again, and it will usually work again.