Last Updated: December 24, 2025
Basic Commands
/imagine prompt: your description here
Generate an image from text prompt
/settings
Open settings menu (select V6 model)
/info
View account info and usage
/subscribe
Subscribe to Midjourney
/prefer suffix
Set default parameters for all prompts
Essential Parameters
| Parameter | Values | Description |
|---|---|---|
--ar |
1:1, 16:9, 9:16, 4:3, 3:2 | Aspect ratio (default 1:1) |
--style raw |
raw | Less opinionated, more literal interpretation |
--stylize |
0-1000 | Artistic interpretation (default 100) |
--chaos |
0-100 | Variation between generations (default 0) |
--weird |
0-3000 | Unconventional aesthetics (V6 feature) |
--quality |
0.25, 0.5, 1 | Render quality/time (default 1) |
--no |
object, color, etc. | Negative prompt (exclude elements) |
--seed |
0-4294967295 | Consistent starting point for generations |
Prompt Structure Best Practices
# V6 Prompt Formula
[Subject] [doing action] [in location/setting], [style/medium],
[lighting], [mood/atmosphere], [camera/composition details]
# Example 1: Photorealistic
a elderly woman reading a book in a cozy library, golden hour
lighting through tall windows, warm and peaceful atmosphere,
shot on Fujifilm XT4, 35mm lens, shallow depth of field --ar 3:2 --style raw
# Example 2: Artistic
a cyberpunk city at night, neon lights reflecting on wet streets,
dystopian mood, digital art, cinematic composition, wide angle --ar 16:9 --stylize 250
# Example 3: Product Photography
luxury watch on marble surface, studio lighting, clean minimalist
background, commercial photography, macro lens, high detail --ar 1:1 --quality 1V6 New Features
Better text rendering
Put text in "quotes" for better accuracy
Improved prompt following
More accurate interpretation of complex prompts
--weird parameter
Create unconventional, unique aesthetics
Natural language
Use complete sentences instead of keywords
Remix mode improvements
Better variation control with remix
Aspect Ratios Guide
| Ratio | Use Case | Example |
|---|---|---|
--ar 1:1 |
Square, social media | Instagram posts, avatars |
--ar 16:9 |
Landscape, widescreen | Desktop wallpapers, YouTube thumbnails |
--ar 9:16 |
Portrait, mobile | Instagram stories, TikTok |
--ar 4:3 |
Classic photography | Prints, traditional photos |
--ar 3:2 |
35mm photography | Professional photography |
--ar 21:9 |
Ultrawide cinematic | Movie posters, banners |
Style Modifiers
# Photography Styles
- "shot on iPhone 15 Pro"
- "35mm film photography"
- "professional product photography"
- "National Geographic style"
- "street photography, Leica M10"
# Art Styles
- "oil painting in the style of Van Gogh"
- "watercolor illustration"
- "digital art, Artstation trending"
- "pixel art, 8-bit style"
- "anime style, Studio Ghibli"
# Rendering Styles
- "photorealistic, 8k, ultra detailed"
- "cinematic lighting, volumetric fog"
- "isometric view, low poly 3D"
- "blueprint, technical drawing"
- "papercraft, layered paper art"Lighting Techniques
golden hour lighting
Warm, soft light during sunrise/sunset
studio lighting
Controlled, professional setup
volumetric lighting
God rays, light beams through atmosphere
rim lighting
Backlight creating edge glow
dramatic lighting
High contrast, moody shadows
soft diffused lighting
Gentle, even illumination
Multi-Prompts & Weights
# Basic Multi-Prompt (separate concepts with ::)
hot::2 dog::1
# Result: Emphasizes "hot" twice as much as "dog"
# Complex Weighted Example
cyberpunk city::3 neon lights::2 rain::1 --ar 16:9
# More emphasis on city, moderate on neon, less on rain
# Negative Weights (avoid elements)
beautiful landscape:: tropical island::2 --no people::1 buildings::1
# Bracket Grouping
{woman reading a book}::2 {cozy library}::1 {fireplace}::1Image Prompts (Image URLs)
# Basic Image Prompt
/imagine https://example.com/image.jpg modern fashion photoshoot --ar 3:2
# Multiple Image References
/imagine https://img1.jpg https://img2.jpg blend these styles --iw 1.5
# Image Weight Parameter (--iw)
https://reference.jpg portrait of a person --iw 2
# Higher values = stronger image influence (default 1, range 0.5-2)
# Combining Image + Text
https://reference.jpg same composition but cyberpunk style --ar 16:9Permutations for Variations
# Test Multiple Options
/imagine a {red, blue, green} sports car --ar 16:9
# Generates 3 separate images
# Multiple Variables
/imagine a {cat, dog} wearing {glasses, hat} --ar 1:1
# Creates: cat with glasses, cat with hat, dog with glasses, dog with hat
# Mix Parameters
/imagine sunset landscape --ar {16:9, 1:1} --stylize {50, 250}
# Tests different ratios and stylize valuesCommon Photography Terms
| Term | Effect |
|---|---|
| Bokeh | Blurred background, shallow depth of field |
| Macro lens | Extreme close-up details |
| Wide angle | Expansive view, slight distortion |
| Telephoto lens | Compressed perspective, distant subjects |
| Tilt-shift | Miniature effect, selective focus |
| Long exposure | Motion blur, light trails |
Stylize Values Guide
| Value | Effect | Best For |
|---|---|---|
| --stylize 0 | Minimal interpretation | Literal prompts, technical drawings |
| --stylize 50 | Subtle artistic touch | Photorealistic images |
| --stylize 100 | Default balance | General use |
| --stylize 250 | Artistic interpretation | Creative, stylized images |
| --stylize 500-1000 | Very artistic | Abstract, highly stylized art |
Pro Tips for V6
Be specific and detailed
V6 understands complex, natural language better
Use "quotes" for text in images
"SALE" "Welcome" renders more accurately
Try --style raw for realism
More photographic, less artistic interpretation
Combine --weird with low stylize
Unique aesthetics without over-stylization
Use remix mode for iterations
Modify prompts while keeping composition
Save good seeds
React with ✉️ to get seed for reproduction
💡 Pro Tip: V6 excels with natural language. Instead of "woman, red dress, cafe, bokeh", try "a woman wearing a flowing red dress sits at a Parisian cafe, soft afternoon light creating bokeh in the background, shot on Leica M10, 50mm f/1.4". The more context and detail, the better your results!