By Apefx Team•February 27, 2026•9 min read
Some prompts consistently produce stunning results across AI models. Not because of magic keywords, but because they follow principles that AI models respond well to — clear composition, specific lighting, defined style, and rich context. Here are 10 prompts that work every time, with detailed breakdowns of why they work so you can adapt them to your own projects.
Why These Prompts Work
Before diving into the prompts, here are the four principles they all follow:
- Specific subject: Not “a person” but “an elderly Japanese potter with weathered hands”
- Defined lighting: Not “good lighting” but “warm rim light from behind, soft fill from the left”
- Clear style reference: Not “nice looking” but “oil painting in the style of Dutch Golden Age masters”
- Composition guidance: Not “portrait” but “close-up, rule of thirds, shallow depth of field, 85mm lens”
AI models were trained on images with rich descriptions. The more your prompt sounds like a professional photographer or art director describing a shot, the better the output.
#1: The Cinematic Portrait
“Cinematic portrait of an elderly Japanese potter with deep wrinkles and kind eyes, holding a freshly thrown ceramic bowl. Warm rim light from a nearby kiln, soft smoke wisps in the background. Shot on 85mm f/1.4, shallow depth of field, film grain, Kodak Portra 400 color palette. Photorealistic.”
Recommended model: Nano Banana Pro (15 credits) or BitDance (4 credits)
Why it works:
- “Elderly Japanese potter with deep wrinkles and kind eyes” — specific character details give the model rich visual information to work with
- “Warm rim light from a nearby kiln” — motivated lighting (light that has a logical source in the scene) looks more natural than generic “good lighting”
- “85mm f/1.4, shallow depth of field” — photography-specific terms that AI models understand precisely. The 85mm focal length produces flattering portrait compression
- “Kodak Portra 400” — referencing a specific film stock tells the model exactly what color palette, contrast, and grain to aim for
#2: The Epic Landscape
“Vast Icelandic highland landscape at golden hour, a single tiny figure standing on a moss-covered ridge overlooking a glacial valley. Dramatic cloud formations with god rays breaking through. Shot on Hasselblad medium format, 50mm, f/8 for maximum depth of field. Scale and solitude.”
Recommended model: Nano Banana Pro (15 credits) or Flux Pro (5 credits)
Why it works:
- “Single tiny figure” — contrast between vast landscape and small human creates instant dramatic scale
- “God rays breaking through” — volumetric lighting is one of the most visually dramatic effects AI models can render
- “Hasselblad medium format” — implies ultra-high resolution and the distinctive medium format look (wider field of view with compressed depth)
- “Scale and solitude” — abstract emotional keywords help guide the overall mood
#3: The Moody Still Life
“Dark moody still life of old books, a brass compass, and a half-melted candle on a worn oak desk. Single light source from an unseen window to the left, deep shadows, dust particles visible in the light beam. Dutch Golden Age painting style, oil on canvas texture, rich earth tones.”
Recommended model: Grok Imagine (7 credits) or Nano Banana Pro (15 credits)
Why it works:
- “Single light source from an unseen window” — chiaroscuro lighting (strong contrast between light and dark) creates drama and depth
- “Dust particles visible in the light beam” — atmospheric detail that makes the scene feel real and lived-in
- “Dutch Golden Age” — a specific, well-represented art movement in training data. AI models produce beautiful results with this reference
- Specific objects: “Brass compass, half-melted candle, worn oak desk” — material and age details give the model texture and character to work with
#4: The Cyberpunk Street
“Narrow cyberpunk alley in a rain-soaked mega-city at night. Neon signs in Japanese and Korean, holographic advertisements reflected in puddles, steam rising from street vents. A lone figure in a translucent raincoat walks away from camera. Blade Runner meets Ghost in the Shell. Ultra-detailed, 8K.”
Recommended model: Nano Banana 2 (8 credits) or Seedream 5.0 (5 credits)
Why it works:
- “Neon signs in Japanese and Korean” — AI models excel at rendering Asian typography in cyberpunk contexts (heavily represented in training data)
- “Reflected in puddles” — reflections add visual complexity and depth that elevate the image from flat to immersive
- “Blade Runner meets Ghost in the Shell” — referencing specific visual works gives the model a clear target aesthetic
- “Lone figure walks away from camera” — back-facing figures are easier for AI to render convincingly than faces, and they add mystery
#5: The Fantasy Character
“Full-body character design of an elven ranger, silver hair in a loose braid, wearing moss-green leather armor with gold filigree details. Standing in a misty ancient forest, bow slung across their back, one hand resting on a wolf companion. Concept art style, clean lines, neutral background fade, character turnaround sheet feel.”
Recommended model: Recraft V4 Pro (8 credits) or Nano Banana Pro (15 credits)
Why it works:
- “Full-body character design” — frames the generation as a design task, encouraging full figure rendering
- “Moss-green leather armor with gold filigree details” — specific material and color descriptions give the model clear rendering targets
- “Concept art style, clean lines” — art direction that produces professional, usable output rather than just “pretty pictures”
- “Character turnaround sheet feel” — a term from concept art that AI models understand, producing a presentation-ready character design
#6: The Architectural Interior
“Minimalist Japanese ryokan interior, tatami mat floors, shoji screen doors, a single low table with a ceramic tea set. Late afternoon sunlight casting long geometric shadows through paper screens. Wabi-sabi aesthetic, imperfect beauty, natural materials. Wide-angle architectural photography, 24mm lens.”
Recommended model: BitDance (4 credits) or Flux Pro (5 credits)
Why it works:
- “Wabi-sabi aesthetic” — references a specific Japanese philosophy that AI models interpret as intentional imperfection, natural textures, and aged beauty
- “Geometric shadows through paper screens” — shadow patterns add visual interest and demonstrate the model’s understanding of light physics
- “24mm lens” — wide-angle focal length gives the sense of space that architectural photography requires
- Minimal but specific: The scene is simple but every element is described with purpose — no clutter, but enough detail to be interesting
#7: The Macro Nature Shot
“Extreme macro photograph of morning dew drops on a spider web, each droplet acting as a tiny lens refracting the sunrise colors behind it. Iridescent web strands against a soft bokeh background of wildflowers. Focus stacking, extreme detail, National Geographic quality.”
Recommended model: Nano Banana Pro (15 credits) or BitDance (4 credits)
Why it works:
- “Each droplet acting as a tiny lens” — describes a specific optical phenomenon that produces visually stunning results
- “Iridescent” — a powerful color keyword that produces rainbow-like light effects
- “Focus stacking” — a real photography technique that produces impossible depth of field in macro, and AI models render it convincingly
- “National Geographic quality” — a quality anchor that references a well-known visual standard
#8: The Vintage Film Photograph
“1970s Kodachrome photograph of a young couple at a seaside carnival at dusk. Cotton candy, ferris wheel lights, motion blur from spinning rides. Warm amber tones, heavy film grain, slight light leak on the left edge. Nostalgic, candid, unposed. Shot on Nikon FM2, 50mm f/1.8.”
Recommended model: Seedream 5.0 (5 credits) or Flux Pro (5 credits)
Why it works:
- “1970s Kodachrome” — era + film stock creates an instantly recognizable visual style: warm, saturated, with that distinctive Kodachrome color rendering
- “Light leak on the left edge” — intentional imperfections make AI images feel authentic rather than sterile
- “Candid, unposed” — guides the model toward natural-feeling compositions rather than stiff, staged setups
- “Nikon FM2” — referencing a specific vintage camera reinforces the era and aesthetic
#9: The Surrealist Composition
“A massive ancient stone staircase spiraling upward through clouds, with doors of different architectural styles floating at each landing — a red Chinese moon gate, a Gothic cathedral door, a modern glass door, a hobbit-hole entrance. A small figure climbs with a lantern. Salvador Dalí meets Hayao Miyazaki, dreamlike atmosphere, soft pastel palette.”
Recommended model: Grok Imagine (7 credits) or Nano Banana Pro (15 credits)
Why it works:
- Impossible architecture: Surreal spatial concepts are something AI excels at — it doesn’t have real-world physics constraints
- Specific details in a surreal context: “Chinese moon gate, Gothic cathedral door, hobbit-hole” — concrete references within an abstract concept
- “Salvador Dalí meets Hayao Miyazaki” — combining two well-known visual artists creates an unexpected but coherent hybrid aesthetic
- “Small figure with a lantern” — provides scale and a focal point for the eye to anchor on in a complex composition
#10: The Product Hero Shot
“Premium product photography of a luxury perfume bottle on a reflective black surface. Dramatic three-point studio lighting — key light from upper right creating a bright highlight on the glass, fill light softening shadows, rim light separating from the dark background. Crystal-clear glass with amber liquid, subtle caustic light patterns. 4K, commercial photography, Hasselblad quality.”
Recommended model: BitDance (4 credits) or Recraft V4 Pro (8 credits)
Why it works:
- “Three-point studio lighting” — the fundamental professional lighting setup, described with specific roles for each light
- “Caustic light patterns” — the light refraction through glass is technically challenging and visually striking. AI models handle it well
- “Reflective black surface” — creates elegant reflections beneath the product, a classic luxury photography technique
- “Commercial photography, Hasselblad quality” — quality anchors that push the model toward professional-grade output
These 10 prompts demonstrate the principles that make AI art generation predictable and high-quality: specific subjects, defined lighting, clear style references, and composition guidance. Adapt these templates to your own subjects and styles. For more on getting started, see our complete beginner’s guide, or jump straight into the AI Art Generator to try them yourself.
Try these prompts right now
50 free credits/month. Copy any prompt above and generate in seconds.
Start Creating Free →