Midjourney V7 is a significant leap forward — better prompt adherence, improved text in images, character consistency, and new parameters. This guide covers everything you need to get excellent results.

Getting Access

Midjourney V7 is available to all paid subscribers. Use it:

  • In Discord: /imagine command
  • On the web: midjourney.com (no Discord needed)

To explicitly use V7: --v 7 at the end of your prompt (V7 is default for most users now).

V7 Key Improvements

Prompt adherence: V7 follows prompts more literally. “A red car parked on a wet street at night” gives you exactly that, not an interpretation.

Text in images: Still not perfect, but significantly better. Short text (4-6 words) usually renders correctly.

Character consistency: With --cref (character reference), maintain consistent characters across generations.

Photorealism: Faces, textures, and lighting are more convincing than V6.

Basic Prompt Structure

[Subject] [Scene/Setting] [Style/Aesthetic] [Lighting] [Camera/Composition] --parameters

Example:

A middle-aged woman scientist in a modern lab examining samples, warm afternoon light through windows, photorealistic, Sony A7 50mm lens

Parameter Reference

Core Parameters

--ar [ratio] — Aspect ratio

--ar 16:9    # Landscape (YouTube, desktop)
--ar 9:16    # Portrait (Instagram Stories, TikTok)
--ar 1:1     # Square (Instagram posts)
--ar 4:3     # Traditional photo
--ar 3:2     # Standard photo

--q [0.25-2] — Quality (default 1)

--q 0.5    # Faster, lower quality (good for drafts)
--q 2      # Maximum quality (2x slower, 2x credits)

--s [0-1000] — Stylize (default 100)

--s 0       # Very literal, raw interpretation
--s 100     # Default Midjourney aesthetic
--s 750     # Strong artistic style, less literal
--s 1000    # Maximum artistic interpretation

--chaos [0-100] — Variation between results

--chaos 0    # Very consistent results
--chaos 50   # Moderate variation
--chaos 100  # Very different results each time

--no [element] — Exclude from image

--no text, watermark, blurry backgrounds

Style References (--sref)

Apply the visual style of a reference image to new generations:

A cozy coffee shop interior --sref [image URL] --sw 100

--sw (style weight) controls how much the style is applied:

  • --sw 0 — no style influence
  • --sw 100 — moderate style influence (default)
  • --sw 1000 — maximum style influence

Example workflow:

  1. Generate a style you love
  2. Save that image URL
  3. Use --sref [URL] to maintain that style across multiple generations

Character References (--cref)

Maintain consistent characters across images:

The same woman from the reference image, now in a park, reading a book --cref [character image URL]

--cw (character weight):

  • --cw 0 — only face matches
  • --cw 100 — full character consistency (clothing, etc.)

Best practices for --cref:

  • Start with a well-lit, front-facing reference
  • Simple, neutral backgrounds work better
  • More references = more consistent results

Effective Prompting Techniques

Describe Light Carefully

Light is the single biggest quality lever:

# Flat, generic lighting (avoid)
"A portrait of a woman"

# Specific, beautiful lighting (do this)
"A portrait of a woman, golden hour light streaming from the left, 
soft shadows on the right side of her face"

Lighting terms that work well:

  • Golden hour / magic hour light
  • Studio three-point lighting
  • Soft diffused natural light
  • Dramatic side lighting / Rembrandt lighting
  • Cinematic blue hour

Camera and Lens Details

For photorealistic images:

Shot on Sony A7R IV, 85mm portrait lens, f/1.8, shallow depth of field
Nikon Z9, 24mm wide angle, architectural photography
Hasselblad medium format, product photography, white seamless

For cinematic images:

35mm film, grain, Kodak Portra 400
Cinematic anamorphic lens flare, widescreen
Shot on RED camera, cinematic color grade

Negative Prompts

What to commonly exclude:

--no ugly, deformed, blurry, watermark, text, extra fingers, bad anatomy

Professional Use Cases

Marketing and Brand Images

[Product] on a clean white surface, product photography, 
soft shadows, commercial quality, professional lighting --ar 1:1 --q 2

Portrait Photography

Professional headshot of [description], wearing [clothing], 
[background], soft studio lighting, Canon 5D Mark IV 
portrait style --ar 4:5

Architecture and Real Estate

Modern Scandinavian living room interior, large floor-to-ceiling windows, 
natural materials, warm afternoon light, architectural photography, 
wide angle lens --ar 16:9

Social Media Graphics

Abstract colorful background, vibrant gradients, modern design, 
suitable for social media header --ar 9:16

Upscaling and Downloading

After generating:

  1. Click image to view full size
  2. U1-U4 — Upscale specific variant (2x resolution)
  3. Vary (Subtle) — Minor variations
  4. Vary (Strong) — Major variations
  5. Zoom Out — Extend the image’s canvas
  6. Download by clicking the image and saving

Credits and Billing

Monthly credits:

  • Basic ($10/mo): 200 image generations
  • Standard ($30/mo): 15 GPU hours (~900 standard images)
  • Pro ($60/mo): 30 GPU hours

Credit consumption:

  • Standard generation: ~1 credit
  • --q 2: ~2 credits
  • Upscaling: ~2 credits
  • Vary Strong: ~1 credit

Use --q 0.5 during exploration to conserve credits; --q 2 for final images only.