How to Write Midjourney Prompts in 2026 (The 7-Pattern Framework)

After generating 500+ Midjourney images for client work in 2026, I can tell you the difference between a good image and a great one is almost never the model - it is the prompt. Midjourney v7 is the most capable image model in the world for aesthetic quality, but it cannot read your mind. The 7 patterns below are the ones that consistently produce publishable output, with 20 real before/after examples to show the difference.

Pattern 1: Subject + Action + Setting + Style + Lighting + Parameters

The structure: Every good prompt has these 6 layers, in this order. The mistake most people make: they write a vibe ("a magical forest") and hope Midjourney fills in the rest. It will - but the result is random. The fix: be specific about each layer.

Weak prompt: "A woman in a forest"

Strong prompt: "A 30-year-old woman with short auburn hair, walking through a dense redwood forest in autumn morning fog, wearing a dark green wool coat, eye-level medium shot, cinematic color grading, 35mm film, soft golden hour light filtering through the trees --ar 3:2 --style raw --v 7"

The 6 layers explained:

• Subject - who or what is in the image (be specific: age, appearance, clothing)
• Action - what they are doing (walking, looking, holding)
• Setting - where they are (indoor/outdoor, location, time of day)
• Style - the aesthetic (photographic, illustrated, painterly, anime)
• Lighting - the light source and quality (golden hour, overcast, neon, studio)
• Parameters - aspect ratio, style version, stylize, etc. (--ar, --style, --v, --s)

Pattern 2: Reference an artist, not a style

Why it works: Saying "in the style of Gregory Crewdson" gives Midjourney 100+ data points of what you mean (creepy suburban nighttime, color palette, composition). Saying "cinematic and moody" gives it 0 data points. Reference artists are 10x more reliable than style keywords.

Before: "A moody portrait of a young woman, cinematic, dramatic lighting"

After: "Portrait of a young woman in a dim motel room, lit by a single bare bulb, by Gregory Crewdson, medium format film, late 1990s aesthetic, 35mm --ar 4:5 --style raw"

Other reliable artist references by category:

• Portrait photography - Annie Leibovitz, Peter Lindbergh, Richard Avedon, Platon
• Fashion photography - Tim Walker, Paolo Roversi, Helmut Newton
• Landscape - Ansel Adams, Edward Burtynsky, Michael Kenna
• Product photography - Karl Blossfeldt, Irving Penn still lifes
• Editorial illustration - Brian Wildsmith, Charley Harper, Jean Jullien

Pattern 3: Camera and lens specify the look

Why it works: "Photorealistic" is too vague. "Canon 35mm f/1.4 shot at f/2.0" tells Midjourney exactly the depth of field, the grain, the color science. The camera-and-lens trick is the difference between "looks like a photo" and "looks like a real photo."

Before: "A product shot of a perfume bottle on a marble table"

After: "Editorial product shot of a luxury perfume bottle on Carrara marble, soft window light from the left, Canon 100mm macro lens, f/2.8, shallow depth of field, soft reflections, advertising photography --ar 4:5 --style raw"

Common lens cheat sheet:

• 24mm - wide, environmental, real estate
• 35mm - street, documentary, editorial
• 50mm - "natural" perspective, portraits, lifestyle
• 85mm - classic portrait, beautiful bokeh
• 100mm macro - product, detail, jewelry, food
• 200mm - sports, wildlife, compressed backgrounds

Pattern 4: Use "by [artist]" + "in the style of [genre]" together

Why it works: Single references are limiting. The combination gives Midjourney two sources of truth. The artist reference pulls the visual language, the genre reference pulls the structure.

Before: "A still life with flowers and fruit"

After: "Still life with peonies, pomegranates, and figs on a dark wooden table, in the style of Dutch Golden Age painting, by Rachel Ruysch, dramatic chiaroscuro lighting, oil painting texture --ar 4:5 --v 7"

Common genre references: Dutch Golden Age, Renaissance, Baroque, Art Nouveau, Bauhaus, Pop Art, Hyperrealism, Brutalist, Wabi-Sabi, Y2K, 1970s Polaroid

Pattern 5: Negative prompts via --no

Why it works: Telling Midjourney what you do NOT want is often as important as telling it what you do want. The --no parameter is the official way. It works.

Before: "A clean modern logo for a coffee shop, simple"

After: "Minimalist logo for a coffee shop, geometric mark, single color, vector style, centered composition, white background --no text, letters, words, characters, people, hands, faces, gradients, drop shadows, 3D, photorealistic, photographic --ar 1:1 --style raw"

Common --no values: text, letters, words, blurry, low quality, distorted, deformed, watermark, signature, nsfw, nude, navel, cropped, out of frame

Pattern 6: Use weights with :: for emphasis

Why it works: When you have a long prompt with multiple subjects, some will be more important than others. The ::N syntax tells Midjourney the relative weight. Default weight is 1, higher numbers mean more emphasis.

Before: "A cat sitting on a book, with a coffee cup next to it, on a wooden table"

After: "A calico cat::2 sitting on a leather-bound book::1.5, with a steaming coffee cup::1 next to it, on a rustic oak table::1, soft window light, 35mm --ar 4:5"

Pro tip: Use weights when you have 3+ subjects and one is the main focus. Do not over-weight (::5 or higher) - the model starts ignoring the rest.

Pattern 7: Iterate, do not start over

Why it works: The biggest mistake in Midjourney is starting a new prompt every time. The right workflow: generate 4 images, pick the closest, use "Vary (Subtle)" or "Vary (Strong)" to evolve it. Use "Upscale" to get the final image. Use "Pan" or "Zoom Out" to extend it.

The right iteration loop:

1. Write the initial prompt (use the 6-layer structure)
2. Generate 4 images, look at all of them
3. Pick the closest to what you want
4. Click "Vary (Subtle)" to get 4 variations
5. Repeat 2-3 times until you have a winner
6. Click "Upscale" for the final 2048x2048 image
7. Use Photoshop or Lightroom to fix color/exposure if needed

The "Vary" buttons are the most underused feature. Most users re-roll the prompt when they should be evolving an image they almost liked.

5 prompt templates you can copy today

Template 1: LinkedIn headshot. "Professional headshot of a [age] year-old [gender] with [hair] hair, wearing a [clothing], neutral expression, soft studio lighting with a single key light from camera left, shot on Canon 85mm f/1.4 at f/2.8, blurred light gray background, corporate photography --ar 4:5 --style raw --v 7"

Template 2: Instagram product shot. "Editorial product photograph of [product] on [surface], soft window light from camera right, shot on Canon 100mm macro at f/4, shallow depth of field, soft reflections, advertising photography --ar 1:1 --style raw --v 7"

Template 3: Blog post header. "Wide editorial photograph of [concept], [mood] lighting, by [artist], [camera] lens, [color palette], cinematic color grading --ar 16:9 --style raw --v 7"

Template 4: Social media quote graphic. "Minimalist still life with [objects] on a plain [color] background, soft directional light, by [artist], no text --ar 1:1 --no text letters words --style raw --v 7"

Template 5: Brand hero image. "Abstract [texture] photograph of [material], by [artist], [color palette], soft directional light, by [photographer] --ar 16:9 --style raw --s 200 --v 7"

What is new in Midjourney v7

If you have not used Midjourney since v5, here is what changed in v7:

• --style raw - more photorealistic, less "AI looking". Use it for product photography and editorial.
• --s (stylize) - range now 0-1000 (was 0-600). Higher = more aesthetic, lower = more literal.
• Multi-image prompts (--cref) - reference images for character/style consistency. Great for series.
• --tile - generates seamless patterns for backgrounds and textures.
• Zoom out and pan - extend images in any direction, not just upscale.

What to do next

For the broader image generation landscape, see our image generators comparison and the full image generation directory. For the prompt patterns in general, see our 9-pattern prompt guide.

Have a Midjourney pattern that works for you? Send it to us and we will feature the best ones in a future update.