Introduction
AI image generation has evolved from a novelty into a powerful tool for designers, marketers, and content creators. Midjourney, DALL‑E 3, and Stable Diffusion each interpret prompts differently, so mastering the art of prompt engineering is essential. This collection of 15 carefully crafted prompts covers styles, composition, and reference techniques. Each prompt includes the exact input, the platform it works best on, the problem it solves, and a description of the result. Use these as templates to jump‑start your own creations.
1. Photorealistic Portrait with Natural Light
Platform: Midjourney v6 | Problem: Generic portraits lack depth. Solution: Add lighting keywords and facial details.
Prompt: Professional headshot, 35mm film, soft window light, freckles, subtle smile, shallow depth of field, hyperrealistic skin texture, natural makeup, award‑winning portrait photographer style --ar 2:3
Result: A studio‑quality portrait with realistic skin pores, catchlights in the eyes, and a creamy bokeh background. The freckles and texture avoid the typical “plastic” look.
2. Cinematic Landscape with Mood
Platform: Stable Diffusion (with Realistic Vision) | Problem: Landscapes feel flat. Solution: Add film grain, camera model, and color grading.
Prompt: Cinematic landscape, foggy ancient redwood forest, morning light piercing through trees, Kodak Portra 400 film grain, anamorphic lens flare, muted teal and orange color palette, epic scale, misty atmosphere, 8k, National Geographic photography
Negative prompt: cartoon, painting, oversaturated
Result: A moody, immersive forest scene with volumetric lighting and a film‑like texture. The negative prompt prevents the AI from drifting into painterly styles.
3. Anime Character in Studio Ghibli Style
Platform: Midjourney v6 | Problem: Generic fantasy characters. Solution: Reference a specific animation studio.
Prompt: Young girl with umbrella standing in a rain‑soaked meadow, Studio Ghibli art style, hand‑painted clouds, soft pastel colors, green grass, glowing fireflies, whimsical, highly detailed, Miyazaki inspired --ar 16:9 --iw 1.5
Result: A frame from an imaginary Ghibli film. The grass is painterly, the colors are soft, and the girl’s expression is gentle. Using --iw 1.5 increases the weight of the style reference from the built‑in Ghibli style.
4. Product Photography on White Background
Platform: DALL‑E 3 | Problem: Inconsistent lighting and shadows. Solution: Use specific lighting terms and environment.
Prompt: Minimalist product photography of a matte black smartphone, placed on a white acrylic surface, studio softbox lighting from top left, soft shadows, reflection on surface, ultra sharp focus, commercial photography, 4k, no text
Result: A clean, magazine‑ready product shot. The smartphone has realistic reflections, and the shadow is soft and natural, exactly what e‑commerce needs.
5. Macro Photography of a Water Droplet
Platform: Stable Diffusion (SDXL) | Problem: Details get muddled. Solution: Use “macro” and describe liquid dynamics.
Prompt: Extreme macro photography, a single water droplet falling into a shallow pool, frozen motion, splash crown, rainbow refraction, seed on the water surface, Canon MP‑E 65mm macro lens, f/2.8, 1/8000 shutter speed, sharp, natural lighting
Negative prompt: blurry, out of focus, painting
Result: A stunning macro shot with a crisp splash and colorful refraction. The high shutter speed keyword freezes the droplet.
6. Isometric Room Interior
Platform: Midjourney v6 | Problem: Isometric perspectives are hard to get right. Solution: Explicit angle and projection type.
Prompt: Isometric bedroom with wooden bed, potted plants, and a laptop on a desk, warm sunlight from a window, top‑down view, orthographic projection, pixel‑art clean style, pastel colors, cozy atmosphere --ar 3:2
Result: A perfect isometric room with equal‑angle walls. The orthographic projection keyword helps Midjourney avoid perspective distortion.
7. Cyberpunk City at Night
Platform: DALL‑E 3 | Problem: Generic neon. Solution: Add architectural details and weather. Prompt: Cyberpunk street in Tokyo, 2077, rain‑soaked asphalt, neon signs in Japanese, holographic advertisements, electric blue and magenta lighting, flying car in distance, crowded sidewalk, wet reflections, Blade Runner vibe, cinematic, 8k
Result: DALL‑E excels at combining many elements. The image shows detailed signs, reflections, and a dark moody sky, with the flying car blurred to suggest motion.
8. Watercolor Painting of a Cityscape
Platform: Stable Diffusion (with Watercolor LoRA) | Problem: AI watercolor looks digital. Solution: Use a LoRA that adds paper texture and brush strokes. Prompt: Watercolor painting of Paris at sunrise, Eiffel tower in the background, row of trees along Seine river, loose brush strokes, visible paper grain, paint splatters, impressionist style, soft colors, wet‑on‑wet effect Negative prompt: photograph, digital art, sharp lines
Result: A genuine‑looking watercolor with paper texture, painted feel, and irregular edges. The LoRA ensures the brush strokes are authentic.
9. Fantasy Creature Design
Platform: Midjourney v6 | Problem: Generic monsters. Solution: Combine anatomy terms and texture keywords. Prompt: Ethereal fox with crystalline antlers, translucent fur, glowing blue veins, digital painting by Simon Stålenhag, bioluminescent forest background, deep frosty blue and cyan palette, high contrast, detailed scales --ar 2:3
Result: A unique creature that looks biologically consistent. The antlers have a crystalline structure with light refractions, and the background complements the creature.
10. Animated GIF Storyboard Frame
Platform: DALL‑E 3 (then animate with other tools) | Problem: Static images. Solution: Design a keyframe. Prompt: Storyboard frame for a 2D animated short, cat chasing a red laser dot through a living room, dynamic pose, motion lines, warm golden hour lighting, simple backgrounds, sketchy pencil line art, clean bold outlines
Result: A frame that looks like it was pulled from an animation storyboard, complete with motion blur lines and expressive cat pose.
11. Minimalist Logo Design
Platform: Stable Diffusion (SDXL) | Problem: Overly complex logos. Solution: Use “vector, flat, geometric”. Prompt: Minimalist logo for a coffee shop, simple line art of a coffee cup with a leaf inside, geometric shapes, flat vector, white background, black lines, scalable, professional, no gradient, negative space
Negative prompt: raster, photorealistic, 3d, shadows
Result: A clean vector‑style logo that can be used directly or traced. The negative prompt prevents any 3D shading.
12. Vintage Travel Poster
Platform: Midjourney v6 | Problem: Vintage style looks modern. Solution: Reference historical art movements. Prompt: Travel poster for the Swiss Alps, vintage 1950s style, bold typography, limited color palette (orange, teal, cream), stylized mountains, retro border, text “Visit Switzerland”, lithograph print texture --ar 2:3
Result: A classic mid‑century poster with halftone dots and slightly offset colors, just like old lithographs.
13. Surreal Surrealism Art
Platform: DALL‑E 3 | Problem: Random elements don’t combine well. Solution: Use a clear surrealist logic (melting, scaling). Prompt: Surrealist painting in the style of Salvador Dalí, melting clocks draped over tree branches, barren landscape with long shadows, distorted perspective, warm golden sky, small ants crawling on surfaces, oil paint texture, dreamlike
Result:** DALL‑E handles the iconic melting clocks with proper reflections. The landscape has that unsettling emptiness typical of Dalí.
14. Scientific Illustration
Platform: Stable Diffusion (with Anatomical LoRA) | Problem: Anatomical inaccuracies. Solution: Use precise terms and reference. Prompt: Scientific illustration of a human heart, cross‑section with labeled chambers (left atrium, right ventricle), watercolor style, vintage medical textbook, subtle grid, Latin terminology, realistic proportions, warm sepia background
Negative prompt: cartoon, 3d render, photorealistic
Result: A detailed anatomical drawing with proper labels and a textbook look, useful for educational materials.
15. Black and White Street Photography
Platform: Midjourney v6 | Problem: Black and white often looks like desaturated color. Solution: Use film stock and contrast keywords. Prompt: Black and white street photography, New York City rainy night, neon sign reflections on wet pavement, man in trench coat walking, high contrast grainy, Ilford HP5 plus 400, documentary style, Henri Cartier‑Bresson composition, decisive moment --ar 3:2
Result: A moody monochrome photo with proper grain and tonal range. The decisive moment keyword encourages a dynamic composition.
Conclusion
These 15 prompts demonstrate how precise language, negative prompts, and artistic references can dramatically improve AI‑generated images. Whether you’re creating marketing assets, concept art, or personal projects, adapt these templates to your needs. For further reading, consult the official Midjourney documentation (docs.midjourney.com), OpenAI’s DALL‑E 3 prompt guide (platform.openai.com), and the Stable Diffusion prompt book (stability.ai/prompt-book). Experiment with parameters like --ar, --iw, and negative prompts—they are the difference between a generic image and a masterpiece.
Comments