Midjourney Prompt Basics
A Midjourney prompt is the text description you type after /imagine. The bot reads it, processes it through its image generation model, and returns four image options. Simple enough - but the difference between a vague prompt and a well-crafted one is enormous.
Midjourney reads your prompt from left to right. Words at the beginning carry more weight than words at the end. If you want "photorealistic portrait," put photorealistic near the start. The tool gives more emphasis to earlier terms when deciding how to render the image.
Don't write natural language sentences. Midjourney responds better to descriptive phrases and keywords. "A woman standing in a forest" is less effective than "woman, standing in forest, dappled sunlight, photorealistic, 85mm lens, shallow depth of field."
Did you know? Midjourney v6.1 achieves near-photorealistic quality at 1024x1024. Adding camera specifications like lens type and f-stop to your prompt dramatically improves realism in portrait and scene photography.
Source: Midjourney documentation and community testing, 2025
Prompt Structure Formula
Here's the formula that works across almost every image type:
[Subject] + [Action/Pose] + [Environment/Background] + [Lighting] + [Style/Medium] + [Mood] + [Technical specs] + [Parameters]
You don't need all eight elements every time. But each one you add gives Midjourney more information to work with and pushes it toward what you actually want.
Let's take a simple subject and build it up:
- Basic: "a coffee shop"
- + Environment: "a cozy coffee shop in Paris"
- + Lighting: "a cozy coffee shop in Paris, warm afternoon light coming through windows"
- + Style: "a cozy coffee shop in Paris, warm afternoon light coming through windows, cinematic photography"
- + Technical: "a cozy coffee shop in Paris, warm afternoon light coming through windows, cinematic photography, 35mm film grain, golden hour --ar 16:9"
Each layer adds specificity. The final version will look dramatically better than the first - more mood, more character, more like something you'd see in a travel magazine.
Style and Aesthetic Keywords
These keywords reliably shift the visual style of Midjourney outputs. Use them as modifiers to any prompt.
Photography styles:
photorealistic- looks like a real photographcinematic photography- movie-quality composition and color gradingeditorial photography- magazine-style with clean compositiondocumentary photography- raw, authentic, unposed35mm film- film grain, slightly muted colors
Art and illustration styles:
digital art- clean, modern illustration feeloil painting- thick brushstrokes, rich colorswatercolor illustration- soft washes, paper textureconcept art- professional game/film concept art qualityanime style- Japanese animation aestheticflat design illustration- minimal, geometric, clean
Era and movement styles:
Art Nouveau- organic forms, decorative flourishesBauhaus- geometric, functional, primary colorsvintage 1970s- warm tones, grainy, retro feelcyberpunk- neon, dark, futuristic urban
Tip: Era References Work Better Than Artist Names
Adding "Art Nouveau style" or "1920s illustration style" produces more consistent results than referencing specific artists. Midjourney's response to artist names is unpredictable - era and movement references are more reliable.
Photorealistic Photography Prompts
Photorealism in Midjourney comes from camera specification. The more it sounds like a real camera setup, the more photorealistic the result. Here are tested prompts with notes on what makes them work.
Portrait Photography
professional headshot of a woman in her 30s, natural smile, white background, Sony A7IV, 85mm f/1.4 lens, studio lighting with key light and fill, sharp focus, professional LinkedIn photo --ar 4:5 --v 6
The camera body, lens, aperture, and lighting type all push this toward a clean professional portrait. The aspect ratio is set for LinkedIn.
Landscape Photography
mountain lake at sunrise, reflections in still water, pine forest on shoreline, mist rising from water, golden hour light, landscape photography, Nikon Z9, 24mm wide angle, high dynamic range, ultra detailed --ar 16:9 --v 6
Wide angle lens specification creates that epic landscape feel. HDR and "ultra detailed" push the detail level.
Street Photography
rainy night street scene in Tokyo, reflections on wet pavement, neon signs, lone pedestrian with umbrella, Leica M11, 35mm f/2, available light, decisive moment, street photography, cinematic --ar 3:2 --v 6
Leica and available light suggest a photojournalistic approach. "Decisive moment" is a famous street photography concept that Midjourney understands.
Illustration and Art Prompts
Character Illustration
fantasy warrior woman, full body character design, armor with intricate engraving, determined expression, forest background, concept art, digital painting, dramatic lighting, fantasy RPG game character, detailed, 4K --ar 2:3 --v 6
Whimsical Illustration
a tiny house inside a teacup, cozy interior visible through windows, flowers growing around the cup, morning light, watercolor illustration style, soft colors, children's book illustration, detailed --ar 1:1 --v 6
Editorial Illustration
person overwhelmed by floating social media notifications, metaphorical illustration, flat design style, limited color palette of blue and orange, editorial illustration for magazine, clean and modern --ar 16:9 --v 6
Design Asset Prompts
Background and Texture
abstract geometric background, gradient from deep purple to teal, hexagonal pattern, subtle noise texture, minimal, tech company website background, dark mode --ar 16:9 --v 6 --no text
Product Mockup Background
clean product photography background, white marble surface, soft shadows, professional studio setup, high key lighting, e-commerce product photography backdrop, empty surface ready for product placement --ar 1:1 --v 6
Social Media Hero Image
minimal flat lay composition, coffee cup, notebook, pen, small plant on white background, top down view, clean lifestyle photography, natural morning light, Instagram aesthetic --ar 1:1 --v 6 --no text, watermark, logo
Advanced Parameters
These are the most useful Midjourney parameters. Each one does something specific.
| Parameter | What it does | Common values |
|---|---|---|
--ar |
Sets aspect ratio | 1:1, 16:9, 9:16, 4:5, 3:2 |
--v 6 |
Sets model version | Use 6 for best quality |
--s |
Stylization (0-1000) | Low=literal, High=artistic |
--c |
Chaos (0-100) | Higher=more varied results |
--no |
Negative prompt | --no text, blur, watermark |
--q |
Quality (0.25, 0.5, 1, 2) | 1 is default, 2 is slower but better |
--sref |
Style reference image URL | Match visual style of any image |
--seed |
Sets random seed for reproducibility | Any number - reuse for consistent results |
The --sref parameter is one of the most powerful. You can paste a URL to any image and Midjourney will try to match that visual style in your generated image. This is how professionals create consistent brand aesthetics across multiple generations.
Negative prompts with --no reduce unwanted elements by around 80%. Common uses: --no text for clean backgrounds, --no people for product shots, --no blur for sharp images.
Prompt Library
Copy these prompts directly. They're tested and produce reliable results.
General Purpose (Change Subject)
[SUBJECT], cinematic photography, dramatic lighting, ultra detailed, professional quality, 8K resolution --ar 16:9 --v 6
Business Headshot
professional business headshot, [person description], neutral background, confident expression, corporate photography, studio lighting, sharp focus, LinkedIn profile photo --ar 4:5 --v 6
Product on White
[product] on pure white background, professional product photography, soft shadow, e-commerce listing photo, studio lighting, clean, minimal --ar 1:1 --v 6 --no text, watermark
Abstract Art
abstract fluid art, [color palette], flowing forms, smooth gradients, minimal, contemporary art gallery quality, high resolution --ar 1:1 --v 6 --s 500
Architectural Interior
modern minimalist interior, [room type], natural light from large windows, clean lines, [style: Japandi/Scandinavian/Industrial], interior design photography, architectural photography --ar 16:9 --v 6
Build Your Personal Prompt Library
Save every prompt that produces a result you love. Note the seed number (shown in the image filename) so you can reproduce similar results. Over time, you'll develop a set of personal "base prompts" that consistently work for your specific use cases.