Midjourney Cheat Sheet

Last Updated: November 21, 2025

Essential Commands

/imagine [prompt]
Generate images from text description
/blend [image1] [image2]
Blend 2-5 images together
/describe [image]
Generate 4 prompts from uploaded image
/settings
View and change settings
/info
Check subscription and usage info
/prefer suffix [text]
Add automatic suffix to all prompts
/shorten [prompt]
Analyze and optimize long prompts
/prefer option set [name] [value]
Create custom reusable option

Key Parameters

Parameter Values Description
--ar or --aspect 16:9, 4:3, 1:1, 2:3, etc. Change aspect ratio (default 1:1)
--v [version] 5.2, 6, niji 5, niji 6 Select model version
--stylize or --s 0-1000 (default 100) Artistic interpretation strength
--quality or --q 0.25, 0.5, 1 (default) Rendering quality/time (affects cost)
--chaos or --c 0-100 (default 0) Variation between grid results
--seed 0-4294967295 Consistent starting point for generation
--style raw, expressive, cute, scenic Apply style variations (v6+)
--tile none Generate seamless tileable pattern
--weird or --w 0-3000 Unconventional aesthetics (v6+)
--no [item] any subject Negative prompt (exclude elements)

Image Weight Parameters

--iw [0.5-2]
Image prompt weight (default 1, v6 only)
--cw [0-100]
Character reference weight (v6 only)
--sref [URL]
Style reference image URL
--cref [URL]
Character reference image URL (v6)
--sw [0-1000]
Style reference weight

Prompt Structure Best Practices

Element Examples Purpose
Subject a medieval knight, cyberpunk cityscape Main focus of image
Medium oil painting, 3D render, photograph Art style/format
Environment in a forest, on Mars, underwater Setting/background
Lighting golden hour, dramatic shadows, neon glow Mood and atmosphere
Color vibrant colors, monochrome, pastel palette Color scheme
Mood serene, chaotic, mysterious, joyful Emotional tone
Composition close-up, wide angle, bird's eye view Camera angle/framing
Artist Style in style of Monet, Studio Ghibli aesthetic Artistic reference

Popular Art Styles

photorealistic, hyperrealistic
Ultra-realistic photographs
anime, manga style
Japanese animation aesthetic
oil painting, watercolor
Traditional painting mediums
digital art, concept art
Modern digital illustration
isometric, low poly
3D geometric styles
pixel art, 8-bit, 16-bit
Retro game aesthetics
art nouveau, art deco
Historical art movements
cyberpunk, steampunk, dieselpunk
Science fiction genres
minimalist, abstract
Simplified artistic approaches
gothic, baroque, renaissance
Classical European styles

Lighting Descriptors

golden hour, magic hour
Warm sunset/sunrise lighting
volumetric lighting, god rays
Light beams through atmosphere
studio lighting, soft box
Professional photography setup
neon lighting, bioluminescent
Glowing artificial/natural light
dramatic shadows, chiaroscuro
High contrast light/dark
backlit, rim lighting
Light from behind subject
ambient occlusion, global illumination
Realistic 3D rendering effects

Camera & Lens Terms

Term Effect Use Case
macro lens, macro photography Extreme close-up detail Tiny subjects, textures
wide angle lens, fisheye Broader field of view Landscapes, architecture
telephoto lens Compressed perspective, zoomed Portraits, distant subjects
depth of field, bokeh, f/1.4 Blurred background Subject isolation
tilt-shift Miniature effect Toy-like appearance
long exposure, motion blur Smooth water/light trails Dynamic movement

Quality Boosting Terms

highly detailed, intricate details
Increase level of detail
8k, 4k resolution, ultra HD
High resolution output
professional photography, award winning
Higher quality standards
octane render, unreal engine
Advanced 3D rendering quality
trending on artstation
Contemporary professional art quality
sharp focus, crisp
Enhanced clarity

Upscaling & Variations

U1, U2, U3, U4
Upscale selected image from grid
V1, V2, V3, V4
Create variations of selected image
🔄 Re-roll
Generate new grid with same prompt
Vary (Strong)
Major variations of upscaled image
Vary (Subtle)
Minor variations of upscaled image
Vary (Region)
Edit specific areas (inpainting)
Zoom Out 2x, 1.5x
Expand canvas with AI generation
Pan arrows
Extend image in direction

Common Prompt Mistakes

Mistake Issue Solution
Too vague "a beautiful scene" Add specific details and context
Conflicting terms "dark bright image" Choose consistent descriptors
Over-prompting Paragraph of details Focus on key elements (use /shorten)
Wrong parameters Incompatible version/feature Check version compatibility
No aspect ratio Cropped subjects Specify --ar for composition

Niji Mode (Anime Specialist)

--niji 6
Latest anime/manga model
--style cute
Kawaii aesthetic (niji 6)
--style expressive
Dynamic, mature anime style
--style scenic
Detailed backgrounds/landscapes
💡 Pro Tip: Start with a simple, clear subject description, then add style, lighting, and composition details. Use --seed to maintain consistency across variations. Try /describe on images you like to learn effective prompt structures. Version 6 understands natural language better, so write prompts more conversationally!
← Back to Data Science & ML | Browse all categories | View all cheat sheets