Midjourney workflow essentials: Prompts, parameters, and resource modes

Once your Midjourney account is set up, it’s time to create your first image. In this section, we’ll walk through how to generate images using both the web interface and Discord. You’ll learn how to enter prompts, select image size and layout, choose a model version, and manage generation speed using Fast, Relax, or Turbo mode. We’ll start with the web interface before covering the Discord workflow—each with clear step-by-step guidance.

This section focuses on the essential generation flow. We’ll return to style settings and advanced parameters in the next section.

Generating your first image on the web interface

The web interface offers a clean and intuitive way to generate images. It’s especially helpful for new users because it allows you to create visuals without memorizing any commands. Here’s how to get started:

Step 1: Go to the Create panel

Log into midjourney.com/app, and click the “Create” tab in the left-hand sidebar. This will open the prompt panel—the main workspace where you’ll build and submit image ideas.

Step 2: Write your image prompt

At the top of the panel, you’ll see a text field labeled “What will you imagine?” This is where you describe the image you want Midjourney to create.

You can type something simple like:

When you open the control panel, you’ll see several settings available on the web interface, including Image Size, Model, Aesthetics, and More Options (Speed). In this section, we’ll focus on Image Size, Model, and Speed. The Aesthetics settings will be covered later.

Step 3: Choose the image size

The Image Size controls the aspect ratio—or shape—of your generated image. Midjourney now offers more options to suit a variety of compositions, all accessible through the layout icons in the web interface.

• Portrait (2:3, 3:4, 9:16, etc.): Best for full-body characters, vertical objects, or mobile wallpapers.
• Square (1:1): A balanced format common for social media and central subjects.
• Landscape (3:2, 4:3, 16:9 etc): Ideal for wide scenes like landscapes, interiors, or cinematic compositions.

Click your preferred icon to set the image’s shape—Midjourney automatically applies the correct --ar value behind the scenes. If you later choose to use aspect ratios via Discord, you’ll need the --ar parameter there (we’ll explain it in detail in the Discord section).

Step 4: Select the model version

The Model setting lets you choose which version of Midjourney’s engine will generate your image. With Version 7 now available, you have even more options and improved flexibility:

• Version 7 (latest): The most advanced release with enhanced coherence, texture quality, and faster performance
• Version 6.1 (default): Still a solid choice, offering reliable detail and prompt accuracy.
• Niji 6: Tailored for anime-style and illustrative artwork.

You’ll also see a Style Mode toggle (Raw on/off), which reduces artistic bias and gives outputs that more literally reflect your prompt. We’ll cover Style Mode and other aesthetic options in their own section later—don’t worry, it’s coming up.

Step 5: Choose a speed mode

The Speed setting affects how quickly your image is generated and how many GPU minutes (credits) it consumes. Choose the one that best fits your workflow:

• Fast Mode: The default—balances speed and credit usage.
• Relax Mode: Slower processing but doesn’t use GPU minutes. Available only to Standard and Pro subscribers.
• Turbo Mode: Generates images fastest but at a higher credit cost.

You’ll find these options at the bottom-right of the Create panel. Click the one that suits your current goal—if you’re just starting out, Fast Mode is a great default choice.

Step 6: Generate and download the image

Once your prompt and settings are ready, click the paper plane button in the top-right corner of the prompt field to submit your image request. Midjourney will begin processing, and within seconds, a grid of four image variations will appear in your feed below.

To download an image, simply right-click on the result and select “Save image” from the dropdown menu. This allows you to save the image directly to your device, just like you would in most browsers.

Generating your first image on Discord

If you're using your own Discord server with the Midjourney bot added, you can generate images by typing into a text channel where the bot is active. This setup gives you more flexibility and control—especially helpful if you want to fine-tune image settings or incorporate Midjourney into a personal workflow.

Unlike the web interface, which relies on visual controls, the Discord experience is built around a command system. In Discord, a command is a typed instruction that begins with a forward slash (/) and tells the bot what to do.

Midjourney supports several commands—for generating images, changing model settings, checking account usage, and switching speed modes. Once you’re familiar with the basic structure, these commands become a fast and efficient way to interact with the bot.

Step 1: Go to a bot-enabled channel

Open a channel where the Midjourney bot has been added and has permission to respond. This can be any text channel in your server, such as #ai-art, #image-lab, or a custom one you created.

Step 2: Use the /imagine command

In the message input bar, type /imagine. Discord will show the /imagine command prompt with a field labeled prompt. This is where you type your image description.

For example:

Once you've entered your description, press Enter to submit. The Midjourney bot will begin generating your image based on the prompt.

Step 3: Add parameters for more control

Just like on the web interface, you can customize key settings by adding parameters directly to your prompt. These optional flags allow you to adjust image layout, model version, and generation speed.In Discord, you can customize how your image is generated by including parameters in your prompt. A parameter is a special keyword that starts with two dashes (--) and tells Midjourney to apply a specific setting—like image size, model version, or generation speed.

These parameters are added after your main prompt. For example:

Here’s what each part means:

• --ar 16:9 sets the aspect ratio to landscape.
• --v 7 selects Version 7, the latest and most capable model.
• --relax tells Midjourney to generate the image in Relax mode, which doesn’t consume GPU credits but takes longer to render. This is ideal when you’re not in a hurry and want to save usage time.

If you don’t add any parameters, Midjourney will use default values:

• Aspect ratio: 1:1 (square format)
• Model version: The last version you used, or Version 6.1 by default
• Speed mode: Whatever mode is currently active on your account
  (You can switch modes manually using /fast, /relax, or /turbo)

We’ll explore more parameters later—including style, references, and blending—but for now, these basics are enough to start experimenting with different results.

Once submitted, the bot will return a grid of four image variations.

Step 4: Generate and download the image

After you submit your prompt, the Midjourney bot will generate a grid of four image variations. Once the result appears in the chat, you can click the image to view it in full size.

To download the image:

• On the desktop, right-click and select Save image as.
• On mobile, tap and hold the image, then choose Save to Photos or the equivalent option.

---

You’ve now completed the essential steps for generating your first image in Midjourney. By understanding how prompts, models, aspect ratios, and speed modes work together, you’re well-equipped to begin experimenting and refining your creative ideas.