How to Use Midjourney: Beginner's Complete Guide 2026

TL;DR / Key Takeaways

– Midjourney no longer requires Discord — you can use it entirely through the web interface at midjourney.com, which I recommend for beginners

– There is no free trial in 2026. Plans start at $10/month (Basic) or $30/month (Standard, recommended)

– Every prompt can use parameters like --ar for aspect ratio, --v for model version, and --style for stylization — learning these five core parameters will immediately improve your results

– Midjourney V8 is now in alpha with dramatically better prompt adherence — it understands detailed instructions far better than V7

– The single biggest mistake beginners make: prompts that are too vague. Specific, detailed prompts consistently outperform generic ones

– For professional work, you need Pro plan ($60/month) to keep images private with Stealth mode

When I started with Midjourney, I spent the first two weeks generating mediocre results and blaming the tool. What I eventually figured out is that Midjourney rewards people who learn how to communicate with it — and that is a learnable skill, not a talent you either have or do not.

This guide is everything I wish I had been told on day one. I have structured it so you can follow along sequentially, from creating your account through to advanced techniques that most users never discover.

[INTERNAL: link to “best AI image generators 2026”]

Part 1: Getting Started With Midjourney

Step 1: Choose Your Plan

Midjourney offers four subscription tiers. There is no free trial in 2026.

Plan	Monthly Price	Annual Price	Fast GPU Hours	Relax Mode	Stealth Mode
Basic	$10	$96 ($8/mo)	3.3 hrs (~200 images)	No	No
Standard	$30	$288 ($24/mo)	15 hrs (~900 images)	Yes (unlimited)	No
Pro	$60	$576 ($48/mo)	30 hrs (~1,800 images)	Yes (unlimited)	Yes
Mega	$120	$1,152 ($96/mo)	60 hrs (~3,600 images)	Yes (unlimited)	Yes

My honest recommendation for beginners: start with Standard ($30/month). Here is why:

The Basic plan’s 200 images per month sounds like a lot until you start generating. When you are exploring and iterating, you will burn through 200 images in a few days. Standard’s unlimited Relax mode changes everything — you can generate as many images as you want at slightly slower speeds (typically 1-5 minutes versus ~30 seconds in Fast mode). For learning and exploration, the unlimited Relax mode on Standard is essential.

If you genuinely want to test Midjourney before committing to Standard, Basic at $10/month is the only option — but manage your expectations on the image count.

Important for companies: If your business earns over $1,000,000 in annual revenue, Midjourney requires you to use the Pro or Mega plan for any commercial use of generated images. This is in the Terms of Service.

[AFFILIATE LINK: Midjourney]

Step 2: Create Your Account

Midjourney now offers two sign-in options:

You no longer need a Discord account to use Midjourney. The web interface at midjourney.com is fully functional and, in my experience, a better starting point than Discord for beginners.

To create your account:

Go to midjourney.com
Click “Sign In” and choose either Google or Discord
Select your subscription plan and complete payment
You will be taken directly to the Create page

That is it. You can start generating immediately after subscribing.

Step 3: Understand the Interface

The Midjourney web interface has three main areas:

Create page: Where you type prompts and see your generation feed. The “Imagine Bar” at the top is where all prompts go.

Organize page: Your personal gallery with folder organization, bulk actions, search by prompt, and filtering by date and parameters.

Editor: An integrated canvas for refining images — inpainting, panning, zooming, outpainting, and selective regeneration.

For Discord users: you can still use Midjourney via Discord if you prefer. The web interface and Discord use the same underlying models and produce identical quality. The difference is workflow: the web interface has visual controls, while Discord uses text commands. Both platforms sync your generation history automatically.

Part 2: Writing Your First Prompt

The Structure of a Good Midjourney Prompt

A Midjourney prompt has a simple structure:

[description of what you want] –[parameter1] [value] –[parameter2] [value]

The parameters always go at the end, after your descriptive text.

Here is an example of a weak prompt versus a strong one:

Weak: a dog

Strong: golden retriever puppy lying in autumn leaves, sunlit forest background, shallow depth of field, Canon 50mm lens, warm afternoon light --ar 3:2 --style raw

The strong prompt specifies: the subject (golden retriever puppy), the action (lying in autumn leaves), the setting (sunlit forest), the lighting (warm afternoon light), the visual style (shallow depth of field, camera lens reference), and includes parameters for aspect ratio and style mode.

You do not need to follow a rigid template, but the principle holds: the more specific your description, the closer Midjourney gets to what you actually have in mind.

What Midjourney Is Good At Understanding

Through extensive testing, I have found Midjourney responds well to:

Subject + action: “a woman reading a book,” “a robot playing chess”
Setting and time of day: “in a rainy Tokyo alley, night, neon reflections”
Art style references: “in the style of a 1970s travel poster,” “Studio Ghibli watercolor style”
Lighting descriptions: “golden hour light,” “dramatic backlighting,” “soft studio light”
Camera and lens references: “shot on 35mm film,” “wide angle lens,” “shallow depth of field”
Mood and atmosphere: “melancholic,” “serene,” “chaotic energy”
Color palette instructions: “muted earth tones,” “vivid neon,” “monochromatic blue”

What Midjourney handles less reliably:

Specific text in images (use Ideogram instead for text-heavy work)
Exact spatial positioning (“person on the left holding a red bag on the right”)
Counting specific objects (“exactly five trees”)
Precise facial features for specific real people (this has intentional restrictions)

Part 3: The Essential Parameters

Parameters are the most powerful tool in your Midjourney toolkit. Beginners who skip this section consistently produce worse results than they could. Here are the ones you need to know.

–ar (Aspect Ratio)

What it does: Sets the dimensions of your generated image.

Syntax: --ar width:height

Common values:

--ar 1:1 — Square (default, social media)
--ar 16:9 — Landscape widescreen (YouTube thumbnails, desktop wallpapers)
--ar 9:16 — Portrait (Instagram Stories, TikTok, phone wallpapers)
--ar 3:2 — Traditional photo aspect ratio (print, website headers)
--ar 4:3 — Standard horizontal (older photo format, presentations)
--ar 2:3 — Portrait photo (book covers, posters)

Practical tip: Always use --ar unless you want a square image. The aspect ratio profoundly affects composition — a 9:16 portrait framing forces Midjourney to compose the image vertically, which changes what ends up in the image.

Example: misty mountain landscape at dawn, pine forests, watercolor style --ar 16:9

–v (Model Version)

What it does: Selects which Midjourney model version generates your image.

Syntax: --v [version number]

Current versions in 2026:

--v 8 — Latest (V8 Alpha, in testing as of March 2026). Dramatically better prompt adherence, improved text rendering, faster generation
--v 7 — Current stable model. Excellent quality, consistent results
--v 6.1 — Previous generation, slightly faster than V7

When to use older versions: Different model versions have different aesthetic characteristics. V6 has a more filmic, slightly grittier aesthetic that some users prefer for certain styles. Do not assume newer always means better for your specific use case — experiment.

Note on V8: As of March 2026, V8 is in alpha on alpha.midjourney.com. Premium features (HD mode, q4, style references) cost 4x more GPU hours in V8 alpha. For standard work, V7 remains the reliable default.

–style (Style Mode)

What it does: Controls how much Midjourney applies its own aesthetic interpretations to your prompt.

Key style options:

--style raw — Minimal stylistic interpretation. Midjourney follows your prompt more literally without adding its own artistic flavor. Best for photorealistic images, product shots, and precise prompt following
--style cute — Adds a soft, cute aesthetic (great for children’s content)
--style expressive — Bolder, more interpretive (artistic work)
--style scenic — Emphasizes atmospheric, landscape-oriented qualities
--style original — Default aesthetic

My most-used combination: For photorealistic images, I almost always add --style raw. It removes Midjourney’s tendency to add painterly or artistic treatments to prompts that call for realism.

Example: product photo of a glass perfume bottle, white background, professional studio lighting --ar 1:1 --style raw

–stylize (Stylization Strength)

What it does: Controls how strongly Midjourney applies its own aesthetic to your image. Think of it as a dial between “follows your prompt exactly” (low values) and “interprets your prompt artistically” (high values).

Syntax: --stylize [value] or --s [value]

Range: 0 to 1000

Default: 100

Practical guide:

--s 0 to --s 50 — Maximum prompt adherence, minimal Midjourney aesthetic. Use when accuracy matters more than beauty
--s 100 — Default. Balanced
--s 300 to --s 600 — Midjourney starts adding artistic interpretation and beauty
--s 750 to --s 1000 — Highly stylized, beautiful but potentially deviating significantly from your prompt description

My recommendation: Start at the default (100) and increase stylize values when you want more visually striking results and you care less about exact prompt accuracy.

–chaos (Variation Level)

What it does: Controls how much variety appears across the four images in your initial generation grid. Low chaos produces four similar images; high chaos produces four very different interpretations.

Syntax: --chaos [value] or --c [value]

Range: 0 to 100

Default: 0

When to use chaos:

--c 0 — When you have a precise vision and want consistent output
--c 25-50 — When you want to explore variations on a theme
--c 75-100 — When you want unexpected, experimental results and are open to surprises

I use --c 20-30 frequently when starting a new project. It generates four meaningfully different interpretations of my prompt and often reveals an angle I had not considered.

–no (Negative Prompting)

What it does: Tells Midjourney what to exclude from your image.

Syntax: --no [element], [element]

Example: professional headshot portrait --no background clutter, shadows, wrinkles, bad lighting

The --no parameter is underused by beginners. When you find that Midjourney keeps adding elements you do not want — extra people in a scene, particular colors, specific objects — use --no to explicitly exclude them.

–q (Quality)

What it does: Controls how much processing time Midjourney spends on your image. Higher quality = more detailed but slower and uses more GPU time.

Syntax: --q [value]

Values:

--q .25 — Faster, more experimental. Good for quick ideation
--q .5 — Faster generation
--q 1 — Default
--q 4 — Available in V8 only. Maximum coherence, 4x slower, 4x more GPU hours

Practical advice: Use --q .25 or --q .5 when rapidly iterating and exploring ideas. Switch to --q 1 (default) or V8’s --q 4 when you are on a final image you want to refine.

Part 4: Discord Setup (For Users Who Want It)

While the web interface is my recommendation for beginners, some users prefer Discord for its community features and rapid keyboard-driven workflow. Here is the setup process.

Setting Up Discord for Midjourney

Step 1: Create a Discord account

Go to discord.com and create a free account. You will need to verify both your email address and phone number to access all Discord features.

Step 2: Join the Midjourney Discord server

Click the “+” button in the left sidebar of Discord (Add a server), then “Join a server.” Paste this link: discord.gg/midjourney

This gives you access to Midjourney’s announcement channels, community galleries, and support channels. You will not generate images directly in this server for long — the public channels move too fast.

Step 3: Create your own private Discord server

Click “+” in the left sidebar, then “Create my own server.” Give it a name. This becomes your private workspace where your Midjourney generations are organized and visible only to you.

Step 4: Invite the Midjourney Bot

In the main Midjourney Discord server, find “Midjourney Bot” in the member list. Click its name, then “Add App” > “Add to server” > select your private server > Authorize.

Step 5: Generate your first image

In any text channel in your private server, type:

/imagine prompt: your description here

Hit enter. The bot will generate a grid of four images in approximately 30-60 seconds (Fast mode) or 1-10 minutes (Relax mode).

Discord Commands You Need to Know

Command	What It Does
`/imagine prompt: [text]`	Generate a new image
`/blend`	Blend 2-5 images together
`/describe`	Upload an image and get prompt suggestions
`/settings`	Open your Midjourney settings panel
`/info`	View your subscription, GPU time remaining

V and U buttons: After a grid generates, you see buttons labeled U1-U4 and V1-V4. U = Upscale (select and enlarge one image). V = Variation (generate variations of one image).

Part 5: Tips for Better Results

Use the Web Editor for Refinement

Once you have an image you like, the web editor at midjourney.com lets you:

Vary Region (inpainting): Select a specific area of an image and regenerate just that part with a new prompt
Pan: Extend the image in any direction to reveal more of the scene
Zoom Out: Expand the image canvas outward to add surrounding context
Remix: Modify your original prompt on an existing image to iterate toward your vision

These tools let you fix specific issues — an odd hand, a cluttered background, the wrong expression — without starting over from scratch.

Build Prompt Templates for Consistency

If you regularly produce images in a consistent style, build a template. Here is an example I use for social media content:

[subject], [action/context], bold geometric flat design, limited color palette of [3 colors], clean white background, icon-style illustration –ar 1:1 –style raw –s 200

Templates save time and ensure visual consistency across a series of images.

Use Style References (`--sref`)

Style references let you feed Midjourney an image whose visual style you want to replicate. Instead of describing the style in words (which is imprecise), you provide a URL to an image.

Syntax: --sref [image URL]

Example: Upload one of your existing designs to a file host, get the URL, then use --sref [URL] to make new images match that visual style. This is a game-changer for brand consistency.

Note: In V8 alpha, --sref costs 4x more GPU hours. In V7, it uses standard GPU time.

Use `--seed` to Reproduce Results

If you generate an image you love and want to create variations that maintain its overall composition, use the seed number.

Every generated image has a seed (a random number that initializes the generation). If you use the same seed with a slightly modified prompt, you get similar images. Find the seed by reacting to your image in Discord with the ✉️ emoji, or on the web by clicking the share icon and viewing image details.

Syntax: --seed [number]

Iterative Refinement Strategy

The workflow that consistently produces my best results:

Start with --c 25 to see multiple interpretations
Pick the best direction from the four images
Upscale (U) the chosen image
Use Vary Region or Remix to fix specific elements
Final refinement in editor for cropping and last adjustments

Resist the temptation to declare an image “done” after the first generation. The best results almost always come from iteration.

Part 6: Advanced Techniques

Character References (`--cref`)

The character reference parameter (--cref) lets you maintain a consistent character appearance across multiple images. Feed it an image of your character, and Midjourney will apply that character’s visual design to new scenes.

This is invaluable for:

Illustrated series with recurring characters
Brand mascots
Children’s book illustrations
Game character art

Syntax: --cref [image URL]

Personalization Profiles (`--p`)

Once you have used Midjourney for a while and rated images, you can create a personalization profile that reflects your aesthetic preferences. Adding --p or --profile to your prompt applies this profile, making Midjourney produce images that lean toward your taste.

Access personalization profiles from your settings page. You can create multiple profiles for different aesthetic needs (e.g., one for client work, one for personal projects).

Moodboard References

On the web interface, you can build moodboards — curated collections of images that collectively communicate a visual direction. When generating, reference your moodboard to guide the aesthetic. This is particularly useful for client presentations and brand work.

Multi-Prompt Syntax

You can give Midjourney multiple separate concepts by separating them with :: followed by a weight number. Higher numbers get more influence over the final image.

Example: forest::2 ancient ruins::1 --ar 16:9

This generates an image that is primarily forest (weight 2) with ancient ruins as a secondary element (weight 1). Negative weights push elements away: forest::1 modern buildings::-0.5 will try to avoid modern buildings while emphasizing the forest.

Part 7: The Midjourney V8 Update (What Changed in March 2026)

Midjourney released V8 to alpha testing in March 2026. Here is what actually matters:

What is genuinely better in V8:

Dramatically improved prompt adherence — complex, multi-element descriptions get followed much more faithfully
Better text rendering within images (though Ideogram still wins for text-heavy work)
Native 2K resolution via --hd mode
Approximately 5x faster standard generation compared to V7
New --q 4 mode for maximum coherence

What to know before switching:

V8 is still in alpha — expect rough edges and evolving behavior
Premium features (HD, q4, style references, moodboards) cost 4x more GPU hours in V8
Some V7 prompts that worked well will produce different results in V8 — prompting style needs adjustment
Relax mode is not yet available in V8 alpha — all V8 generations use Fast GPU hours

My recommendation: Use V7 as your default while experimenting with V8 on new projects. When V8 exits alpha, it will likely become the new standard. Keep an eye on the Midjourney updates page at updates.midjourney.com for the full release.

Part 8: Pricing Decisions — Which Plan Is Right for You?

Based on the usage patterns I have observed (and my own), here is how to decide:

Choose Basic ($10/month) if:

You genuinely generate fewer than 200 images per month
You want to test whether Midjourney is right for you before committing to Standard

Choose Standard ($30/month) if:

You generate images for personal projects, content creation, or freelance work
You want unlimited generation via Relax mode
Your business earns under $1M in annual revenue

Choose Pro ($60/month) if:

Your business earns over $1M in annual revenue (required for commercial licensing)
You need Stealth mode to keep client work private from Midjourney’s public gallery
You need 12 concurrent Fast jobs for high-throughput workflows

Choose Mega ($120/month) if:

You generate 100+ images per day regularly
You run an AI image agency or production studio

All plans get a 20% discount on annual billing. If you know you will use Midjourney for a year, annual billing reduces Standard to $24/month effective — a meaningful saving.

Want to Go Deeper? The Midjourney Prompt Bible

If you want to dramatically accelerate your prompt-writing skills, a comprehensive prompt guide walks through advanced composition techniques, style vocabulary, lighting terminology, and 200+ tested prompt templates across 12 categories.

Whether you are creating for social media, commercial photography, character art, or illustration, having a reference library of proven prompt structures saves hours of trial and error. A digital prompt guide can pay for itself in the first week of use.

[INTERNAL: link to “Midjourney prompt guide product”]

FAQ: How to Use Midjourney

Do you still need Discord to use Midjourney in 2026?

No. Midjourney’s web interface at midjourney.com is now fully functional and does not require Discord. You can sign in with Google. Discord remains available and has some advantages for community learning and advanced command workflows, but beginners will find the web interface much easier to start with.

How long does it take for Midjourney to generate an image?

In Fast mode, standard generations typically complete in 20-60 seconds depending on server load. In Relax mode (available on Standard, Pro, and Mega plans), generation times range from 1-10 minutes. V8 alpha is approximately 5x faster than V7 in standard mode. Premium features like --hd and --q 4 take 4x longer to generate.

What is the best Midjourney prompt formula for beginners?

Start with this structure: [main subject] + [action or pose] + [setting] + [lighting] + [art style] + [camera/lens if photorealistic] + --ar [ratio]. For example: “a woman writing in a notebook, cozy coffee shop interior, warm afternoon light, Canon 85mm lens, shallow depth of field –ar 3:2”. The more specific each element, the better the results.

How do I make Midjourney generate the same image style consistently?

Use --seed [number] to reproduce similar compositions with modified prompts. Use --sref [image URL] to apply a specific visual style from a reference image. Build a personalization profile after generating and rating 200+ images. For character consistency, use --cref [image URL] to maintain a character’s appearance.

What is the difference between the Midjourney web interface and Discord?

Both platforms produce identical image quality using the same AI models. The web interface offers a visual editor with integrated inpainting, panning, and organization tools — better for beginners and visual-first workflows. Discord offers text-command speed for power users and community features for learning from other artists. Most experienced users use both, generating on Discord and organizing on the web interface.

Midjourney pricing and features verified as of March 2026. V8 is currently in alpha — features and costs may change as it moves toward full release. Check updates.midjourney.com for the latest.