How to Use Midjourney AI: A Complete Guide (Updated)

Midjourney is an AI image generator that turns text descriptions and visual references into images.

You can use it to create illustrations, product concepts, portraits, landscapes, marketing visuals, character designs, and other creative assets.

The easiest way to get started is through Midjourney’s website. Its Create page includes a prompt bar, image-generation feed, reference-image controls, and tools for refining your results.

You can also use Midjourney through Discord if you prefer a command-based workflow.

The steps below cover account setup, prompt writing, image generation, editing, advanced settings, pricing, and practical tips for getting better results.

What Is Midjourney AI?

Midjourney is an AI-powered image-generation platform that creates visuals from text prompts and image references. You describe the subject, style, composition, lighting, or other details, and the model generates images based on your instructions.

For example, a prompt such as:

A small wooden cabin beside a lake, surrounded by pine trees, early morning mist, soft natural light, cinematic landscape photography

can produce several interpretations of that scene.

Midjourney is used for:

• Digital art and illustration: Create concept art, fantasy scenes, character designs, and stylized artwork.
• Marketing and advertising: Develop campaign concepts, promotional visuals, and social media graphics.
• Product visualization: Explore packaging, product concepts, and creative directions.
• Editorial content: Generate illustrations and visual assets for articles, presentations, and publications.
• Creative projects: Experiment with photography styles, architecture, interiors, and visual storytelling.

Midjourney is not a conventional photo editor or a general-purpose design application. Its main purpose is generating and refining images, although its website also includes tools for editing existing visuals.

What Do You Need Before Using Midjourney?

Before generating images, you need a Midjourney account and an active subscription for regular image creation.

Here is what you should prepare:

1. A Midjourney account: Register or sign in through the official website.
2. A subscription: Choose a plan that provides the GPU time and features you need.
3. A text prompt: Decide what you want the image to show. You can start with a simple description.
4. Optional reference images: Upload an image if you want to guide the subject, visual style, or composition.
5. A web browser or Discord account: The website is the simpler starting point, while Discord is an alternative interface.

Midjourney does not currently offer a general free trial on its website or through Discord. A limited trial is available through the Niji Journey mobile app, which is a separate anime-focused experience.

How to Use Midjourney on the Website?

The Midjourney website is the most straightforward way for beginners to generate and manage images.

Its Create page brings the prompt bar, generation feed, and image controls together in one interface.

Step 1: Create a Midjourney Account

Visit midjourney.com and sign in using one of the available login methods.

After signing in:

1. Open your Midjourney account.
2. Review the available subscription plans.
3. Select a plan and complete the subscription process.
4. Open the Create page from the website navigation.

You can use the same subscription on the website and Discord. You do not need to purchase separate subscriptions for the two interfaces.

Step 2: Open the Create Page

The Create page is where you enter prompts and generate images.

At the top, you will find the Imagine bar, which is the main input area for describing your image. Depending on the interface and settings available to your account, you can also access image uploads, personalization controls, and generation settings from this area.

The central creation feed displays your generated images. You can return to previous creations, open individual images, and continue editing or generating variations.

Step 3: Write Your First Prompt

Click the Imagine bar and enter a description of the image you want to create.

For example:

A modern glass house in a forest, surrounded by tall pine trees, morning sunlight, architectural photography

Press Enter or click the send button to submit the prompt.

Midjourney will process the request and display the generated results in your creation feed.

Tip: Start with a clear description rather than a long list of instructions. Add more details only when they help communicate the subject, composition, or visual style.

Step 4: Review the Generated Images

Once generation finishes, open the results and compare the different interpretations.

Look for the image that best matches your intent:

• Subject and setting
• Composition and camera angle
• Lighting and color palette
• Artistic or photographic style
• Overall mood

You do not need to accept the first result. Midjourney is designed for experimentation, so you can adjust the prompt or use the available image actions to explore alternatives.

Step 5: Refine Your Selected Image

After choosing an image, use the available creation actions to continue working on it.

Depending on the model and image, you can use tools such as:

• Variations: Generate alternative interpretations of an image.
• Upscale: Prepare a selected result at a larger output size using available upscaling options.
• Pan: Extend the image beyond its original boundaries.
• Zoom Out: Expand the visible scene around the original image.
• Editor: Make more targeted changes to an image, including changing selected areas or expanding the canvas.

The available actions can vary depending on the model used to create the image.

Step 6: Download Your Image

When you are satisfied with the result, open the image and use the download option in its image controls.

Choose the available download option and save the file to your device. You can then use it in a design project, presentation, website, or other permitted application.

For professional publishing, review the image before using it. Check for visual inconsistencies, inaccurate details, unwanted objects, and text errors.

How to Use Midjourney in Discord?

Midjourney also supports image generation through Discord.

This method uses slash commands and is useful if you already work in Discord or prefer a command-based interface.

Your Midjourney subscription must be connected to your Discord account before you can generate images there.

Step 1: Connect Your Discord Account

1. Sign in to Midjourney on its website.
2. Open your account settings.
3. Find the Discord connection option.
4. Select Connect Discord.
5. Sign in to the correct Discord account and authorize the connection.

Check the Discord username before authorizing the connection. Using the wrong account can create problems when accessing your subscription.

Step 2: Open a Midjourney Bot Conversation or Server

After connecting your account, open Discord and access a space where the Midjourney Bot is available.

You can use an appropriate Midjourney channel or interact with the bot through a direct message, subject to the available Discord settings and permissions.

Step 3: Enter the /imagine Command

In the Discord message box:

1. Type /.
2. Select the /imagine command from the command list.
3. Enter your image description in the prompt field.
4. Press Enter or Send.

For example:

/imagine prompt: A futuristic city at sunset, flying vehicles, glass skyscrapers, cinematic science fiction concept art

The bot will process your request and return generated image options.

Step 4: Select and Refine an Image

Review the generated results and use the available buttons or commands to continue.

Depending on the interface and model, you may be able to:

• Create variations of a selected image.
• Upscale a result.
• Rerun the original prompt.
• Adjust the prompt and generate another version.

The website and Discord provide overlapping functionality, but their controls are not identical.

The website generally offers a more integrated editing workflow, while Discord relies more heavily on commands and manual prompt management.

Midjourney Website vs. Discord: Which Should You Use?

Both interfaces use your Midjourney subscription, but they suit different workflows.

Recommendation: Start with the website if you are new to Midjourney. Its visual interface makes it easier to manage prompts, reference images, and generated results. You can move to Discord later if its command-based workflow better suits your needs.

How to Write Better Midjourney Prompts?

A prompt is the description you provide to Midjourney to guide image generation. It can be a single phrase or a detailed description of the scene.

The quality of your prompt affects how closely the generated image matches your idea.

However, adding more words does not automatically produce better results. Clear and relevant details are usually more useful than lengthy instructions.

Understand the Main Elements of a Prompt

A useful prompt can include several elements:

You do not need to include every element in every prompt. Choose the details that matter to the image you want.

Midjourney Prompt Examples

Here are a few examples showing how prompts can be structured for different creative tasks.

1. Photorealistic image

A close-up portrait of an elderly fisherman, weathered face, natural skin texture, soft window light, documentary photography

2. Product photography

A premium matte black perfume bottle on a dark stone surface, soft studio lighting, subtle reflections, luxury product photography

3. Landscape illustration

A quiet mountain village surrounded by snow-covered peaks, pine forests, early morning fog, detailed digital illustration

4. Social media visual

Minimalist workspace with a laptop, notebook and coffee cup, warm natural light, clean composition, modern editorial photography --ar 4:5

5. Website hero image

Abstract architectural structure made of translucent glass, soft blue-gray palette, clean negative space, subtle shadows, premium technology editorial visual --ar 16:9

These examples are starting points. Adjust the subject, visual style, and composition to suit your project.

Keep Prompts Focused

Midjourney’s official prompting guidance recommends using clear descriptions and avoiding unnecessary lists of instructions.

For example, this prompt is relatively vague:

A beautiful house

A more useful version is:

A contemporary concrete house surrounded by olive trees, large glass windows, warm afternoon sunlight, architectural photography

The second prompt provides information about the architecture, setting, lighting, and visual medium. That gives Midjourney more context without making the description unnecessarily complicated.

What are the important Midjourney Parameters You Should Know?

Parameters are additional instructions placed at the end of a prompt. They help control image proportions, artistic interpretation, variation, and other generation settings.

Parameters begin with two hyphens and should be separated from the prompt text by a space.

For example:

A modern Japanese garden with a stone bridge and koi pond --ar 16:9

Here are some useful parameters to learn.

1. Aspect Ratio (--ar)

The aspect ratio controls the width-to-height proportions of an image.

Examples:

• --ar 1:1 — Square image.
• --ar 16:9 — Landscape or widescreen image.
• --ar 9:16 — Vertical image.
• --ar 4:5 — Portrait-oriented image.
• --ar 3:2 — Traditional landscape photography proportions.

For example:

A mountain landscape at sunrise --ar 16:9

Use an aspect ratio that matches the intended placement, such as a website banner, social media post, or portrait layout. Aspect ratio does not specify the final pixel dimensions.

2. Stylize (--stylize or --s)

The stylize parameter controls how much of Midjourney’s artistic interpretation influences the result.

• Lower values generally follow the prompt more closely.
• Higher values allow more artistic interpretation.

Example:

A red vintage sports car on a mountain road --stylize 50

The default stylize value in the latest versions is 100, and the parameter can be adjusted from 0 to 1000.

3. Raw Mode (--raw)

Raw Mode reduces some of Midjourney’s automatic stylistic influence, giving you more control over how the prompt is interpreted.

It can be useful when you want a more literal result or are aiming for a specific photographic appearance.

Example:

A professional studio portrait of a woman against a neutral background --raw

Raw Mode is compatible with Midjourney versions 5.1 and later.

4. Chaos (--chaos or --c)

Chaos controls how varied the initial image results are.

A lower value generally produces more similar results, while a higher value encourages greater variation.

Example:

A futuristic city skyline at night --chaos 50

Use it when you want to explore different compositions rather than receive closely related interpretations.

5. No (--no)

The --no parameter tells Midjourney which elements you want to exclude from the image.

Example:

Minimalist product photography of a white ceramic vase --no flowers

It can help reduce unwanted elements, although it does not guarantee that every excluded detail will be absent.

6. Seed (--seed)

A seed is a number used to initialize an image-generation job. It can help with controlled experiments when you want to compare prompt changes.

Example:

A small cabin in a snowy forest --seed 12345

A seed does not guarantee an identical image across different model versions or settings. Treat it as an experimental control rather than a way to lock an exact composition.

Important: Parameters can have different compatibility rules across model versions. Check Midjourney’s current Parameter List before combining advanced controls.

How to Use Images as References in Midjourney?

You do not have to rely only on text prompts. Midjourney supports image-based inputs that can guide the appearance or style of a new creation.

On the website, you can upload an image through the image menu beside the Imagine bar and assign it an available reference type.

1. How to Upload and Use a Reference Image?

To use an image as a reference, open the Imagine bar on the Midjourney website and select the image upload option.

Choose an image from your device, then assign it to the appropriate reference category if that option is available.

Next, enter a text prompt describing the new image you want to create.

For example, you could upload a photograph of a modern building and prompt Midjourney to generate a similar architectural concept with a different setting, materials, or lighting.

Choose the reference type according to your goal.

Use an Image Prompt when you want to influence the generated image’s visual content or composition, a Style Reference when you want to guide its aesthetic, and a Moodboard when you want to apply a broader visual direction.

2. Image Prompts

An Image Prompt uses an existing image to influence the content, composition, or visual direction of a new generation.

For example, you could upload a photograph of a building and prompt Midjourney to create a different architectural scene inspired by its structure.

Image Prompts are useful when a written description alone cannot communicate the visual idea clearly.

3. Style References

A Style Reference helps transfer the visual characteristics of an image, such as its color palette, lighting, texture, and artistic medium.

For example, you could use a watercolor painting as a Style Reference and generate a new landscape with a similar visual treatment.

On the website, select an uploaded image and assign it as a Style Reference. In Discord, you can use the --sref parameter followed by an image URL.

The --sw parameter controls how strongly the Style Reference influences the result.

4. Personalization

Personalization allows Midjourney to adapt its image generation to your aesthetic preferences.

You can select images you like to develop a Personalization profile. Once enabled, the profile can influence future generations.

Personalization is useful when you want a more consistent visual direction across multiple projects without describing every stylistic preference in each prompt.

5. Moodboards

Moodboards let you collect images that represent a particular creative direction. You can use them to guide generations toward a broader aesthetic.

For example, a moodboard for a premium interior-design project might include references showing neutral colors, natural materials, minimalist furniture, and soft lighting.

Moodboards are particularly useful when you want to maintain a visual identity across a series of images.

How to Edit Images in Midjourney?

Generating an image is only one part of the workflow. You can also refine existing images using Midjourney’s website tools.

The Editor is available from the Edit page and can also be opened from certain image actions in your creation feed.

1. How to Edit an Image Step by Step?

To edit an image, open one of your generated creations and select the available Editor option. You can also open the Editor from the website’s editing interface and upload an existing image if supported by your workflow.

Once the image is open, choose the area or part of the canvas you want to change. Use inpainting to modify a selected region, or outpainting to extend the image beyond its original boundaries. You can then enter instructions describing the changes you want Midjourney to make.

Review the generated result and make further adjustments if necessary. When you are satisfied, use the available download option to save the edited image.

Exact placement: Add this H3 inside the existing “How to Edit Images in Midjourney” H2, immediately after the introductory paragraph and before the existing “Use the Midjourney Editor” H3.

This makes the section more useful without creating another H2.

2. Use the Midjourney Editor

The Editor lets you make changes to Midjourney images and upload images from outside the platform.

Its tools include:

• Inpainting: Regenerate or modify selected areas of an image.
• Outpainting: Expand the image beyond its original boundaries.
• Pan: Extend the canvas in a particular direction.
• Zoom Out: Create a wider view around the original composition.
• Remix: Adjust prompt details while creating variations.

The exact editing workflow depends on the model and the tools available for the image.

3. Understand the V8.2 Edit Model

Midjourney introduced its V8.2 Edit Model in 2026. It supports instruction-based editing, image references, inpainting, and outpainting.

The V8.2 Edit Model also replaces older workflows involving Omni Reference, Character Reference, and Retexture in the V8.X experience.

This is an important distinction when following older Midjourney tutorials: some instructions written for earlier versions may refer to controls that have since changed or been replaced.

Which Midjourney Model Should You Use?

Midjourney’s model versions have changed over time, with improvements to image quality, prompt interpretation, and creative controls.

As of November 2026, V8.2 is the default image model. It focuses on aesthetics, image quality, and improved Personalization.

For most new users, the default model is a sensible starting point.

You generally do not need to switch versions unless you have a specific reason, such as reproducing an older workflow or comparing outputs.

Before using a parameter or feature from an older tutorial, check whether it is supported by the model you have selected.

How Much Does Midjourney Cost?

Midjourney is a subscription-based service. Its four main plans are Basic, Standard, Pro, and Mega.

The official pricing currently lists the following subscription rates:

Annual subscriptions are billed upfront and offer a 20% discount compared with paying monthly.

Which Midjourney Plan Should You Choose?

• Basic: Suitable for occasional image generation and users who are still learning the platform.
• Standard: A practical option for regular creators. It includes unlimited image generations in Relax Mode, subject to the plan’s feature conditions.
• Pro: Designed for heavier use. It provides more Fast GPU time, Stealth Mode, and additional video-generation benefits.
• Mega: Intended for users who need the largest allocation of Fast GPU time among the standard plans.

Relax Mode for images is available on Standard, Pro, and Mega plans. It allows unlimited image generations without consuming monthly Fast GPU time, although generation can take longer.

Prices are listed in US dollars. Check the official subscription page for current billing details, taxes, and any changes to plan features.

What Can You Create With Midjourney?

Midjourney can support a range of creative workflows. The best approach depends on the type of output you need and how much refinement the project requires.

1. Marketing and Social Media Graphics

You can generate campaign concepts, promotional illustrations, product backgrounds, and visual ideas for social media.

For example, a marketing team might generate several visual directions for a product launch before selecting one for further design work.

2. Website and Blog Images

Midjourney can create editorial illustrations, website banners, article visuals, and background images.

For a website hero image, specify the subject, visual style, color palette, and aspect ratio. If you need space for a headline, describe the desired negative space in the prompt.

3. Product and Packaging Concepts

Designers can use Midjourney to explore product shapes, packaging concepts, materials, and visual directions.

These outputs are useful for early-stage ideation, but they should not automatically be treated as production-ready technical designs.

4. Concept Art and Creative Projects

Artists and creators can use Midjourney to explore characters, environments, costumes, fictional locations, and different illustration styles.

Image references and Personalization can help maintain a more consistent visual direction across a project.

What are the Common Midjourney Mistakes to Avoid?

New users often focus on adding more prompt details without considering how those details affect the final image. A few practical adjustments can make the process more efficient.

1. Writing vague prompts: Describe the subject and important visual details instead of relying on broad words such as “beautiful” or “professional.”
2. Adding conflicting instructions: Avoid combining styles or requirements that work against each other.
3. Using too many parameters: Learn a few useful controls first rather than adding several settings to every prompt.
4. Ignoring aspect ratio: Choose image proportions based on where the visual will be used.
5. Expecting exact text: Although newer models have improved text rendering, always check spelling, layout, and small text before publishing.
6. Skipping image refinement: Use variations, editing, and upscaling when the first generation is close but not quite right.
7. Following outdated tutorials: Midjourney changes its models and interface, so older instructions may no longer match the current tools.

How to Troubleshoot Common Midjourney Problems?

Suggested content:

If Midjourney does not behave as expected, first check whether the issue is related to your account, prompt, subscription, or interface.

• Image generation does not start: Check your account status, subscription, and available GPU time. Then try submitting the prompt again.
• Your subscription is not recognized: Confirm that you are signed in to the correct Midjourney account. If you use Discord, check that the correct Discord account is connected.
• Discord commands are unavailable: Make sure you are using a space where the Midjourney Bot is available and that you have selected the correct /imagine command.
• The generated image looks different from the prompt: Simplify the description, remove conflicting instructions, and adjust the important visual details.
• An editing option is unavailable: Check whether the selected model and image support that editing action.

If the issue continues, consult Midjourney’s official Help Center for current account, generation, and interface guidance.

What Are Midjourney’s Limitations?

Midjourney can produce detailed and visually compelling images, but it has limitations that matter when using the results in professional projects.

• Exact instructions: The model may interpret a prompt differently from what you intended, particularly when it contains many constraints.
• Text accuracy: Text inside images can still contain spelling errors, distorted letters, or inconsistent typography.
• Fine details: Small objects, hands, repeated patterns, and complex arrangements may need closer inspection or editing.
• Consistency: Maintaining the exact appearance of a character or product across multiple images can require careful use of references and compatible tools.
• Technical precision: Generated product concepts, diagrams, and architectural visuals should not be treated as verified engineering specifications.
• Privacy: Creations are public by default on Midjourney’s platform. Stealth Mode is available only with Pro and Mega plans, and images made in shared Discord spaces can remain visible to other participants.

For commercial work, also review Midjourney’s terms and the relevant copyright rules in your jurisdiction.

Midjourney states that users own the images and videos they create, subject to exceptions in its terms.

For example, upscaling another user’s image does not transfer ownership to you.

Businesses with more than $1 million in annual gross revenue must use a Pro or Mega plan for commercial use under Midjourney’s stated policy.

Conclusion

Learning how to use Midjourney involves more than entering a prompt and downloading an image. The process includes describing your idea clearly, choosing suitable settings, reviewing generated results, and refining the images with variations and editing tools.

For beginners, the website’s Create page is the easiest place to start. Once you understand basic prompting, experiment with aspect ratios, Style References, Personalization, and the Editor. Midjourney’s V8.2 model is the current default as of November 2026, so use its latest documentation when following tutorials that cover older versions.

Frequently Asked Questions (FAQs)

1. Is Midjourney free to use?

Midjourney does not currently offer a general free trial on its website or through Discord. A limited trial is available through the Niji Journey mobile app. Regular image generation requires a paid subscription.

2. How do beginners use Midjourney?

Beginners can start by signing in at midjourney.com, opening the Create page, and entering a simple description in the Imagine bar. After generating images, they can select a result, create variations, refine it, and download the finished image.

3. Can I use Midjourney without Discord?

Yes. You can use Midjourney directly through its website. Discord is an alternative interface, not a requirement for website-based image generation.

4. What is the latest Midjourney version?

As of November 2026, Midjourney V8.2 is the default image model. It focuses on aesthetics, image quality, and improved Personalization.

5. Can I upload my own images to Midjourney?

Yes. You can upload images through the website and use them as Image Prompts, Style References, or supported editing references. The available options depend on the model and workflow.

6. Can I use Midjourney images commercially?

Midjourney states that users own the images and videos they create, subject to the exceptions in its Terms of Service. Businesses with more than $1 million in annual gross revenue need a Pro or Mega plan for commercial use under its stated policy. Copyright laws also vary by country, so review the applicable rules before using generated images commercially.

Also Read –

What is Midjourney? A Comprehensive Guide (2026)

Midjourney V8 Model: Faster AI Image Generation