🎨 Midjourney Mastery Prompts Pack — 30 Prompts from v6 to v7 Aesthetics

Your Midjourney Session — 5-Image Scandi Minimalist Productivity Series, v7

Aesthetic Strategy

For a 5-image cluster with tight aesthetic coherence, --sref (style reference) is the right tool — build ONE strong anchor image first, then use it as style reference for all 4 follow-up images. This is the 2026 workflow that beats 'copy the same prompt 5 times hoping for consistency.'

Your description (Scandi minimalist, muted beige/cream, warm natural light, photographic-not-literal) is specific enough to hit right the first time. We'll build the anchor prompt carefully.

Prompts I'm Generating

Prompt 1: The Anchor Image (establishes aesthetic)

Minimalist Scandinavian workspace, soft morning sunlight streaming through linen curtains, wooden desk with single ceramic coffee mug and open leather journal, muted warm tones of beige cream and soft gray, shallow depth of field, photographic editorial style, analog film aesthetic, natural composition with negative space --ar 16:9 --style raw --stylize 300 --v 7

Breakdown:

• Subject first: 'Minimalist Scandinavian workspace'
• Atmosphere: 'soft morning sunlight streaming through linen curtains'
• Specific props (but simple): 'wooden desk with single ceramic coffee mug and open leather journal' — 2-3 props max for minimalist
• Color palette explicit: 'muted warm tones of beige cream and soft gray'
• Technical: 'shallow depth of field' (softens background), 'photographic editorial style' (magazine look), 'analog film aesthetic' (adds warmth)
• Composition: 'natural composition with negative space' (empty space IS the aesthetic)
• Parameters: 16:9 for blog hero, --style raw for photorealistic not stylized, --stylize 300 for modest artistic liberty, v7 explicit

Expected result: Warm-toned editorial photograph feel. Wooden desk, soft morning light through curtains, 1-2 objects, lots of empty space. If first generation looks right, save the image URL — you need it for the next 4 prompts.

Iteration direction: If too literal/photo-stock-ish, raise --stylize to 500. If too painterly/stylized, lower to 150 and ensure --style raw is present. If too dark, add 'bright morning' to the atmosphere phrase. If too cluttered, remove one prop.

Prompts 2-5: Using --sref for style consistency

Once Prompt 1 gives you the anchor image, copy its URL (click the image in Midjourney → 'Copy link' → that's your --sref value).

Then for the 4 follow-up images:

[your --sref URL] Single hand holding matcha latte in ceramic bowl, morning light, blurred minimalist desk background --ar 16:9 --style raw --sref [URL] --sw 100 --v 7

[your --sref URL] Open notebook with fountain pen on wooden surface, soft shadow, minimalist composition, natural light --ar 16:9 --style raw --sref [URL] --sw 100 --v 7

[your --sref URL] Person reading book in window-side armchair, warm morning glow, scandi interior, quiet atmosphere --ar 16:9 --style raw --sref [URL] --sw 100 --v 7

[your --sref URL] Empty desk with single plant and coffee cup, overhead flat lay, minimalist workspace, soft diffused light --ar 16:9 --style raw --sref [URL] --sw 100 --v 7

Each follow-up:

• Subject shifts (matcha / notebook / reading / empty desk) — the content varies
• Style inheritance via --sref (you paste the URL from Image 1 at both start AND as --sref parameter)
• --sw 100 is 'style weight' — at 100, strong style inheritance from reference
• Same aspect ratio, same --style raw, same version

All 5 images will feel like siblings — same light quality, same color palette, same photographic register. Different subjects, consistent aesthetic.

Reference Mode Usage (--sref workflow)

For v7 specifically, style references work at --sw range 0-1000. Default 100. Higher = more style inheritance but less flexibility on content. For tight cluster like yours: 100 is right. For loose family-feel: 50. For strict 'this must look identical': 200-400 (but content flexibility drops).

The Full 30-Prompt Library (Copy Ready)

CATEGORY 1: Aesthetic Development

1.1 The Mood Board Reference Lock

Generate 4 mood-establishing images with same lighting, palette, register. Pick the strongest. Use as --sref for all future images in that aesthetic. Reusable brand asset.

1.2 The Visual Vocabulary Builder

For your aesthetic, build a list of 10 specific descriptors that appear across prompts (lighting, color, composition terms). Consistency of vocabulary = consistency of output.

1.3 The Style-Mix Exploration

Combine 2-3 aesthetic references (80s Polaroid + Wes Anderson + Hopper painting). Use --chaos 25-50 for creative combinations. Finds unexpected but cohesive hybrid aesthetics.

1.4 The Negative-Space Composition

Explicit 'with generous negative space' / 'wide empty composition' in prompts. Midjourney defaults to filling frame; negative-space language pulls back. Strong for hero images needing text overlay.

1.5 The Personalization (--p in v7)

After 200+ ratings in v7, add --p flag. Midjourney pulls aesthetic from your rating history. Most under-used v7 feature. Ratings are your training data.

CATEGORY 2: Tool-Specific Parameters

2.1 The --style raw for Photorealistic

Default Midjourney has subtle stylization baked in. For commercial / editorial / photorealistic work: always add --style raw. More literal rendering.

2.2 The --stylize Sweet Spot

--s 250 for balanced prompt-literalness + artistic quality. --s 500 for more stylized work. --s 1000 for Midjourney to 'interpret.' --s 50 for very literal. Rarely use defaults.

2.3 The --chaos for Ideation

--chaos 25 for slight variation across 4-grid. --c 50 for more varied outputs. --c 100 for wildly different takes on prompt. Use for brainstorming, not production shots.

2.4 The --weird Creative Surprise

--weird 250-500 for unusual compositions + unexpected aesthetic choices. --w 1000+ for creative-wild. Genuinely surprising outputs; good for ideation.

2.5 The --ar Aspect Ratio Planning

--ar 16:9 (blog/video thumbnail), --ar 9:16 (phone/vertical social), --ar 1:1 (Instagram square), --ar 2:3 (print portrait), --ar 3:2 (print landscape). Check target FIRST.

CATEGORY 3: Character Consistency

3.1 The Character Reference Setup (--cref)

Generate 1 strong image of your character. Copy URL. For every follow-up of same character: include --cref [URL] --cw 50 (character weight). Keeps features consistent.

3.2 The --cw (Character Weight) Tuning

--cw 100 = tight to reference (features locked). --cw 50 = balanced. --cw 0-30 = loose (just 'inspired by'). Start 100 for strict consistency; relax as needed.

3.3 The Multi-Pose Character Series

Same character in different poses: keep --cref constant, vary pose description in prompt text. 'Same [character name/description] standing / sitting / walking / laughing.' Consistent face, varied bodies.

3.4 The Character Wardrobe Changes

For character in different outfits: --cref keeps the face; vary wardrobe description. 'Same [character] wearing business suit / casual weekend / formal gown.' Face inherits; clothes flex.

3.5 The Character + Style Reference Stack

Combine --cref + --sref in same prompt. 'Same character [--cref URL], [scene description], [aesthetic], --sref [style URL] --sw 100 --cref [char URL] --cw 80.' Multi-layer consistency.

CATEGORY 4: Product Photography

4.1 The Studio Setup Prompt

Specific lighting + background + camera specs. 'Product shot on seamless white background, studio softbox lighting from 45 degrees, shallow depth of field, 85mm lens, commercial photography, minimalist composition.' Much more likely to get usable e-commerce output.

4.2 The Material-Specific Prompt

Specify material explicitly: 'matte ceramic mug' / 'polished chrome bottle' / 'distressed leather wallet' / 'sheer cotton fabric.' Materials render differently; specificity matters.

4.3 The Background Control

'On pure white seamless' / 'on warm wood surface' / 'on textured concrete' / 'minimal gradient background.' Don't leave background to Midjourney interpretation.

4.4 The Product + Context Scene

Lifestyle product shot: product IN use-context, not just isolated. 'Ceramic mug on wooden desk with steam rising, morning kitchen setting, natural light from window.' More Instagram-feed-ready.

4.5 The Product Variation Grid

Use --chaos 50 when you want 4 different angles/lighting of same product. Great for A/B selection. Pick strongest, iterate from there.

CATEGORY 5: Stylized Illustration

5.1 The Children's Book Style

'Watercolor children's book illustration style, soft pastel colors, simple composition, child-friendly' + character + scene. Add --stylize 600 for artistic interpretation.

5.2 The Concept Art Style

'Concept art, digital painting, cinematic composition, dramatic lighting, game environment' + specific scene. Strong for fantasy / sci-fi worldbuilding.

5.3 The Editorial Illustration

'Editorial illustration style, limited color palette, bold shapes, magazine illustration' + subject. Strong for blog posts, newsletter covers, conceptual pieces.

5.4 The Comic / Graphic Novel Style

'Comic book illustration, bold lines, halftone shading, dynamic composition, graphic novel style' + scene. Specify inspiration ('in the style of [genre], not [specific artist name]').

5.5 The Isometric Game Asset

'Isometric perspective, clean line art, game asset style, flat colors, minimal shading' + specific object/scene. Good for web design assets, app interfaces, game mockups.

CATEGORY 6: Creative Experimentation

6.1 The --weird Exploration

'[Subject + style] --weird 500 --chaos 25.' Forces Midjourney into unusual interpretations. 2-3 of 4 may be unusable; 1 is often genuinely surprising.

6.2 The Permutation Syntax

'{modernist, brutalist, art deco} architecture' — Midjourney generates one image per variant. Rapid style comparison.

6.3 The Remix Mode Iteration

After first generation, use remix (R) button + rewrite prompt for one of the 4 images. Keeps the strong composition; adjusts details. Faster than starting from scratch.

6.4 The Cross-Genre Prompt

'Modern office space rendered as medieval fantasy illustration' / 'Cyberpunk kitchen in watercolor style.' Genre + subject collision often produces fresh work.

6.5 The Deliberate Minimalism

'Single [object] on empty background, minimal composition, lots of negative space.' Counter to Midjourney's tendency toward visual richness. Strong for focused brand assets.

Troubleshooting

If output feels generic / stock-photo-like:

Your prompt is too vague or too stuffed with common keywords. Strip to core subject + specific aesthetic. Add --style raw + one unique composition detail ('wide empty right side,' 'shot from low angle,' 'harsh shadow across frame').

If consistency is off across series:

You're relying on same-text-prompt instead of --sref. Generate one strong anchor, use --sref for all follow-ups. Text alone can't maintain consistency across generations.

If photorealistic looks AI-generated (plastic skin, weird hands, mirror-symmetric faces):

Add --style raw. Add specific details ('skin texture visible, slight imperfection, natural asymmetry'). Avoid 'ultrarealistic 8k' — counterintuitively often makes output LESS real.

If style drifts across iterations:

You're over-tweaking text between iterations. Fix ONE thing per iteration. Remix mode is your friend — generate + select best + remix one detail.

If product photography looks cartoon-ish:

--style raw missing probably. Add specific lighting setup. Add camera specs (85mm lens, f/2.8). Reference 'commercial product photography' explicitly.

If you're getting same 4 images back:

--chaos 0 (default) produces similar outputs. Add --chaos 25-50 for variation. Also try --weird 200 for creative deviation.

Variation Playbook

Brand / Commercial:

Category 3 (Character Consistency via --cref) + Category 4 (Product Photography) primary. --sref for aesthetic lock across brand assets. Consistency > creativity.

Children's Book / Editorial Illustration:

Category 3 (--cref for character across 30+ scenes) + Category 5.1 (Children's Book Style). Develop one strong character in 4-5 reference images before starting production.

Concept Artist (Games/Film):

Category 1.3 (Style-Mix Exploration) + Category 6 (Experimentation) primary. --chaos + --weird for ideation phase; then --sref for locking aesthetic.

E-Commerce Product Photography:

Category 4 primary. Studio setup prompting + material specificity + background control. --style raw always. Consistent lighting setup across product line.

Beginner (First 30 Days):

Category 2 (Parameters) + Category 1.2 (Visual Vocabulary) primary. Focus on ONE aesthetic for 30 days. Master one thing before expanding.

v7 Personalization Builder:

Rate 200+ images intentionally (not random swiping). Your ratings become training data. After threshold, --p flag in prompts pulls from your aesthetic. Takes 2-4 weeks of rating.

Key Takeaways

• Prompt order matters. Subject → style → setting → lighting → technical → parameters. Earlier words weight heavier. Keyword-stuffing kills coherence.
• Reference modes are the 2026 game-changer. --sref for style consistency, --cref for character consistency. Text alone can't maintain consistency across images.
• --style raw for photorealistic. Default style is subtly stylized; raw is more literal. Most commercial/editorial work needs --style raw.
• --stylize 250-500 is the sweet spot for most work. Default 100 = literal; 1000 = Midjourney artistic liberty. Rarely use defaults.
• Iterate through remix mode. First generation is ideation; production shots come through 3-5 iterations. Fix ONE thing per iteration; don't rewrite wholesale.