How to use Midjourney? Tips and tricks


Midjourney, an artificial intelligence image generation tool, has just launched version 6! Want to develop your skills? We’ve got the best tips for learning how to use Midjourney.

Midjourney is an AI program that generates images from textual descriptions.

  • This tool has just launched its new version: V6, now available to everyone in beta.
  • Among the improvements: better image quality and a finer understanding of prompts.

To test Midjourney v6, users must generate an image from Discord with a prompt that begins with /imagine and must add the mention –v 6.0. This way, Midjourney will generate an image from the latest version before the engine integrates it directly by default.

Want to learn how to use Midjourney? Take advantage of all our advice to improve your skills.

Join the Midjourney Discord

Its use is becoming more widespread, particularly due to the lack of legislation on the use of generated photos and their unique character. The tool, therefore, seems to guarantee the production of images not found elsewhere and of excellent quality.

Midjourney runs on the Discord platform. Discord is an instant messaging software initially used by video game players. You must join the Midjourney Discord server to have access to the platform.

You must create a Discord account (for free) and join the Midjourney server. To find it, simply click on “Explore Communities,” the search bar in the middle of the screen. You must then join a Midjourney “Newbies” room on Discord, indicated on the left of your screen.

Midjourney dallefree

Adjust the Desired Image Settings

Once in one of the “newbies” rooms, you will need to adjust the image generation parameters so that the results proposed by the AI are as close as possible to your search.

  • To access the settings, enter the dialog box and type the command “/settings” then press Enter.
  • You will then access different parameters that you can adjust according to your preferences, such as image quality, style, execution speed, etc.

As you become more comfortable with Midjourney, you can return to these settings to adjust them according to your needs.

Learn to Master Prompt Writing

To make Midjourney work, you must use a prompt. Midjourney is an AI image generator. It bases its work on the data previously provided to it.

To make a request to Midjourney, you must write a prompt corresponding to the instructions you want the software to follow when creating your image. Write your prompt in English because the software has difficulty understanding other languages effectively.

The more details you provide in your prompt, the better you will guide the tool in generating the desired image. Certain points should be highlighted for effective understanding.

For example, Midjourney does not include exclusion. This means that if you ask for “a bicycle without wheels,” it is unlikely that you will get the image you want. Focus on inclusive descriptions.

You can orient the style of images using references to artists. For example, you could request a coastal landscape in the style of painter Pablo Picasso.

Choose the main subject of the image, the environment, the lighting, the angle—in short, everything that could be chosen in a traditional photo.

Small details make a difference, so don’t hesitate to be as precise as possible while remaining concrete.

Finally, in the Discord room, when you create images, type the command “imagine” preceded by a slash: /imagine. Then add your prompt and start creating.

Midjourney ai free

Select the Correct Image

Once your prompt has been processed, Midjourney presents you with 4 images. Below these images, you’ll find 4 U buttons and 4 V buttons. The U buttons allow you to enlarge your desired image among the 4 offered, while the V buttons generate 4 new images based on the one you like the most from the initial set. These commands help you refine your selection.

Once you have chosen and enlarged the photo, you can zoom in or out and adjust the shooting angle to see what is on the sides, top, or bottom of the image. This way, you can experiment with your image, its environment, and the artificial intelligence.

Test and Retest

Image generation is an interesting but time-consuming process. Once you have a specific type of image in mind, it’s unlikely you’ll achieve it on the first try.

You will probably need to use the V command, edit your prompt, rephrase a sentence, add details, or take a different angle on your prompt. To get a specific image, don’t be afraid to spend several minutes or even hours on it.