15 Curated AI Image Prompt Examples
Fifteen ready-to-use prompts covering the most requested scenes in AI image generation: realistic portraits, anime characters, fantasy landscapes, cyberpunk cities, watercolor illustrations, sci-fi concept art, food photography, architectural visualization, oil painting, dark fantasy, slice-of-life, pixel art, 3D renders, and nature photography. Every prompt is hand-written from Promgrammer’s curated tag library, so you can copy it directly or remix it in the builder.
Each example carries a negative prompt, a tag-by-tag breakdown of why each token is there, variation tips, a model recommendation, and the image the prompt actually produced.
Realistic Portrait Photography
A camera-oriented portrait with lifelike skin texture and natural lighting. Suitable for ad portfolios or profile shots. We combine camera vocabulary (aperture, focal length) with lighting terms (rim light, key light) so the model recognizes photographic conventions. Most stable on SDXL/Flux realism checkpoints.
Prompt
masterpiece, best_quality, 8k, photorealistic, 1girl, solo, portrait, upper body, looking at viewer, gentle smile, long black hair, brown eyes, soft skin texture, shallow depth of field, bokeh, studio lighting, rim light, 85mm lens, f/1.8, Kodak Portra 400 film grain
Tap a tag to carry it over to the builder
Negative Prompt
low quality, worst quality, blurry, plastic skin, oversaturated, deformed, extra fingers, watermark, text, logo
Tag Breakdown — Why these tokens?
Global quality baseline — nudges the model to prioritize detail and resolution
Locks the art style to realism, suppressing anime-style noise
Fixes subject count; prevents drift into multi-person compositions
Photographic DoF convention — softens the background via bokeh
Rim highlight separates subject silhouette from background
Evokes film stock characteristics for warm skin tones
Variations
- ▸ Male subject: `1girl, solo` → `1boy, solo, short hair`
- ▸ Outdoor golden hour: `studio lighting, rim light` → `golden hour, warm sunlight, outdoor`
- ▸ Monochrome: append `monochrome, black and white photography`
- ▸ Wider framing: `85mm lens` → `35mm lens, environmental portrait`
Recommended Model
SDXL Realistic / Flux.1-dev
Anime checkpoints tend to produce plasticky skin for this prompt
Related tag categories
Anime Character Design
An anime-styled character sheet with crisp lines and cel shading. Specifying outfit, pose, and expression tightens the character design and performs best on anime-tuned checkpoints like NovelAI, Animagine, or Pony Diffusion. Lock direction early with `1girl, solo, anime style`.
Prompt
masterpiece, best_quality, amazing_quality, anime style, 1girl, solo, full body, long silver hair, blue eyes, school uniform, pleated skirt, standing, peace sign, gentle smile, cherry blossoms, spring, cel shading, vivid colors
Tap a tag to carry it over to the builder
Negative Prompt
low quality, bad anatomy, extra limbs, extra fingers, muscular, realistic, photo, 3d, blurry, watermark
Tag Breakdown — Why these tokens?
Locks the style to 2D anime, suppressing 3D rendering drift
Framing directive — ensures a full character reference pose
Core identity tags following the 3–5 appearance-tag rule
Two or three wardrobe tags to define silhouette clearly
Seasonal backdrop tags that set mood in one stroke
One pose + one expression avoids contradictory gestures
Variations
- ▸ Fantasy variant: `school uniform` → `mage robe, staff in hand, magical aura`
- ▸ Chibi: prepend `chibi, super deformed, kawaii`
- ▸ Dark mood: `cherry blossoms, spring` → `abandoned hallway, dim light, horror atmosphere`
- ▸ Duo shot: `1girl, solo` → `2girls, hugging, looking at each other`
Recommended Model
Animagine XL / NovelAI v3 / Pony Diffusion V6
Realistic checkpoints will not render cel shading correctly
Related tag categories
Fantasy Landscape
A sprawling fantasy landscape emphasizing epic scale. We lean on background and lighting tags while composition cues like `epic scenery`, `wide angle`, and `panoramic` establish scale without a subject. Ideal for game concept art or matte paintings; appending a renderer cue (Unreal Engine, Octane) enriches detail.
Prompt
masterpiece, best_quality, 8k, epic scenery, fantasy landscape, floating islands, ancient ruins, waterfall, giant crystals, misty mountains, aurora borealis, twilight, wide angle, panoramic view, volumetric lighting, god rays, Unreal Engine 5, highly detailed matte painting
Tap a tag to carry it over to the builder
Negative Prompt
people, person, human, character, text, watermark, low quality, blurry, oversaturated, cartoon
Tag Breakdown — Why these tokens?
Stacks three scale cues: framing, focal length, and field of view
Four hero landmarks — anchor points for the viewer’s eye
Setting + phenomenon + time-of-day locks the mood
Rays cutting through mist — the signature fantasy atmosphere
Renderer cues — pushes detail density and crisp tone mapping
Variations
- ▸ Underwater fantasy: `floating islands` → `underwater ruins, bioluminescent coral, deep sea`
- ▸ Snowy peaks: `twilight` → `blizzard, snow-covered peaks, arctic sunrise`
- ▸ Infernal scene: `aurora borealis, twilight` → `lava rivers, black skies, demonic ruins`
- ▸ Zoom-in: `panoramic view` → `close-up of a single ruin, moss-covered stones`
Recommended Model
SDXL Base / Juggernaut XL / Midjourney v6
SDXL-class models preserve resolution best in ultra-wide scenes
Related tag categories
Cyberpunk Cityscape
A classic cyberpunk city drenched in neon and rain. The combination of color tags (cyan, magenta, neon pink) with mood tags (rainy, foggy) is the heart of the look, and a phrase that describes the mood directly (e.g., `neon noir atmosphere`) quickly defines the visual language. In Promgrammer, cross-selecting `colors`, `lighting`, and `theme` reproduces this combo.
Prompt
masterpiece, best_quality, 8k, cyberpunk cityscape, neon signs, rain-soaked streets, flying cars, towering skyscrapers, holographic advertisements, night, foggy atmosphere, cinematic lighting, reflections on wet pavement, cyan and magenta color scheme, neon noir atmosphere, dystopian future
Tap a tag to carry it over to the builder
Negative Prompt
daytime, clear sky, rural, nature, forest, low quality, blurry, oversaturated, watermark
Tag Breakdown — Why these tokens?
Genre + signature element placed up front for strong bias
Wet asphalt doubles the neon color footprint via reflection
Future-tech signals — raises SF detail density
Palette lock — strengthens neon contrast
Compresses the wet-neon night mood into one phrase
Variations
- ▸ Neo-Tokyo: `neon noir atmosphere` → `neo-tokyo, japanese signs, kanji, 1980s (style), retro anime`
- ▸ Daytime cyberpunk: `night, foggy` → `overcast day, drone traffic, overcast cyberpunk daylight`
- ▸ Back alley: `towering skyscrapers` → `narrow back alley, steam vents, cyberpunk noodle stall`
- ▸ Add a character: append `1robot, humanoid android, looking at viewer`
Recommended Model
SDXL Cyberpunk LoRA / Juggernaut XL
Neon renders without a LoRA, but XL-class checkpoints lead on detail
Related tag categories
Chibi Character
A 2–3 head-height deformed character. `chibi, super deformed` compresses proportions while bright palettes and oversized eyes maximize cuteness. Ideal for stickers, profile icons, and emoticon sets; solid or transparent backgrounds maximize reusability.
Prompt
masterpiece, best_quality, chibi, super deformed, 1girl, solo, kawaii, big eyes, pink hair, twin tails, school uniform, red bow, happy expression, waving hand, pastel colors, simple white background, sticker style, thick outline
Tap a tag to carry it over to the builder
Negative Prompt
realistic, photo, detailed face, mature, muscular, nsfw, complex background, blurry, dark, horror
Tag Breakdown — Why these tokens?
Force-compresses proportions; prevents drift to normal anatomy
Enlarges face area and pushes the cute factor
Sticker visual language — bold outline + soft palette
Empty background for easy compositing into other designs
Action + emotion combo, ideal for sticker use cases
Variations
- ▸ Fantasy chibi: `school uniform` → `wizard hat, tiny staff, magical sparkles`
- ▸ Kemonomimi: `1girl` → `1cat, cat ears, cat tail, anthropomorphic`
- ▸ Sad variant: `happy expression, waving hand` → `sad expression, crying, tears`
- ▸ Team set: `solo` → `3girls, chibi trio, group pose`
Recommended Model
Animagine XL + chibi LoRA / Pastel Mix
Base anime checkpoints work, but a dedicated chibi LoRA stabilizes proportions
Related tag categories
Watercolor Illustration
A watercolor illustration mimicking paint bleeding into paper. Combining texture tags (`watercolor`, `paper texture`) with soft lighting intentionally dampens the harsh digital sheen. Great for children’s books, wedding invites, or travel art — leaving deliberate negative space preserves watercolor’s signature breathing room.
Prompt
masterpiece, best_quality, watercolor illustration, paper texture, soft color bleed, 1girl, solo, reading a book, sitting under a tree, autumn leaves, golden hour, delicate brush strokes, pastel palette, loose linework, white space around subject
Tap a tag to carry it over to the builder
Negative Prompt
digital painting, 3d, photo, sharp edges, heavy outline, oversaturated, vector art
Tag Breakdown — Why these tokens?
Declares the medium itself, preventing drift to digital painting
Highlights watercolor-specific bleed and relaxed line quality
Softens brushwork and chroma intensity
Deliberate negative space — watercolor’s signature breathing
Seasonal cue + lighting doubles the warmth
Variations
- ▸ Winter theme: `autumn leaves` → `falling snow, winter coat, steaming mug`
- ▸ Architectural watercolor: `1girl` → `venice canal, gondola, architectural sketch`
- ▸ Animal lead: `1girl, reading a book` → `1cat, sleeping on a windowsill`
- ▸ Rich palette: `pastel palette` → `jewel tones, rich watercolor`
Recommended Model
SDXL + watercolor LoRA / DreamShaper XL
Watercolor effects stabilize most reliably with a LoRA
Related tag categories
Sci-Fi Concept Art
A scene from a sci-fi universe — space station, alien life, or mecha — rendered in concept-art style. Industry terms like `concept art`, `matte painting`, `key frame` summon the correct pre-production look. To raise background detail density, include prop tags like machinery, pipes, and wires.
Prompt
masterpiece, best_quality, 8k, concept art, key frame, sci-fi space station interior, futuristic corridor, glowing panels, holographic displays, 1 astronaut, floating in zero gravity, reflective helmet visor, intricate mechanical details, pipes and wires, cool color temperature, volumetric lighting, cinematic composition, matte painting style
Tap a tag to carry it over to the builder
Negative Prompt
fantasy, medieval, watercolor, cute, chibi, low detail, blurry, watermark, logo
Tag Breakdown — Why these tokens?
Industry lexicon anchors the exact production style
Locates the scene in two layers: place + sub-environment
Raises background density — avoids empty-looking interiors
Locks the palette cool for SF alienation
A neat trick: the visor’s reflection hides face rendering difficulty
Variations
- ▸ Alien planet: `space station interior` → `alien planet surface, twin moons, exotic flora`
- ▸ Mecha scene: `1 astronaut` → `1 mecha pilot cockpit, heads-up display, giant robot silhouette outside`
- ▸ Dystopian: append `abandoned, overgrown plants, dust particles`
- ▸ Warm tone: `cool color temperature` → `warm amber lighting, sunrise through viewport`
Recommended Model
SDXL Base / Juggernaut XL / RevAnimated
Hardware detail is richest on SDXL-class checkpoints
Related tag categories
Food Photography
A food-photography prompt for menus, blogs, or Instagram. In food shots, lighting dictates 90% of the result — invest in cues like `natural window light` or `soft box lighting`, and use food-native framings such as `top-down view` or `45-degree angle`. Steam, smoke, or sauce drizzle injects kinetic appeal.
Prompt
masterpiece, best_quality, 8k, photorealistic, food photography, gourmet dish, juicy steak, grilled vegetables, garlic butter sauce, rustic wooden table, steam rising, natural window light, soft shadows, 45-degree angle, shallow depth of field, rich colors, Michelin-star presentation
Tap a tag to carry it over to the builder
Negative Prompt
cartoon, anime, plastic, artificial, low quality, blurry, oversaturated, person, hands holding plate
Tag Breakdown — Why these tokens?
Genre lock — prevents drift into fine-art illustration
Textbook food-photography lighting setup
Angle that shows both volume and detail of the dish
Visual cue for warmth and freshness
Sets context (homey vs fine dining) explicitly
Variations
- ▸ Dessert: `juicy steak, grilled vegetables` → `chocolate lava cake, vanilla ice cream, berry compote`
- ▸ Asian cuisine: `steak` → `ramen bowl, soft boiled egg, chopsticks`
- ▸ Overhead view: `45-degree angle` → `top-down view, flat lay composition`
- ▸ Beverage add-on: append `glass of red wine, condensation droplets`
Recommended Model
SDXL Realistic / RealVisXL / Juggernaut XL
Anime checkpoints render food as toy-like; not recommended
Related tag categories
Architectural Visualization
An architectural visualization prompt for proposals, portfolios, and real-estate ads. Declare the architectural style (Scandinavian, Brutalist, Art Deco), time of day, and a renderer name (`V-Ray`, `Corona Render`, `Lumion`). If people are needed, use brief scale references like `human scale figures, walking`.
Prompt
masterpiece, best_quality, 8k, architectural visualization, modern scandinavian villa, floor-to-ceiling windows, wooden cladding, minimalist design, forest surroundings, snow on ground, dusk, warm interior lights, cinematic composition, V-Ray render, high detail, long exposure, reflection on pond
Tap a tag to carry it over to the builder
Negative Prompt
cartoon, painting, anime, low detail, deformed structure, people (large), cluttered, blurry, watermark
Tag Breakdown — Why these tokens?
Genre + renderer — two-layer style lock
One-token style declaration compresses the design language
Two structural features that lock the silhouette
Cool exterior + warm interior contrast — archviz cliche for good reason
Adds a symmetry axis, elevating sophistication
Variations
- ▸ Urban apartment: `forest surroundings, snow` → `urban skyline, city lights, rooftop garden`
- ▸ Brutalism: `modern scandinavian villa` → `brutalist concrete building, raw texture`
- ▸ Interior: prepend `interior shot, living room, fireplace`
- ▸ Art Deco: `minimalist design` → `art deco style, gold accents, 1920s glamour`
Recommended Model
SDXL + archviz LoRA / Juggernaut XL
Renderer names act as strong style cues on realistic checkpoints
Related tag categories
Oil Painting Style
An oil-painting look with thick brushwork and tangible texture. Medium-specific tags like `oil painting`, `thick impasto`, and `visible brush strokes` do the heavy lifting, while historical painter references (Rembrandt, Caravaggio) transfer their lighting and composition instincts. Pairs beautifully with `chiaroscuro`.
Prompt
masterpiece, best_quality, classical oil painting, thick impasto, visible brush strokes, 1 woman, renaissance portrait, wearing red velvet dress, holding a letter, candlelight, chiaroscuro, dramatic shadows, rembrandt lighting, warm earth tones, museum quality, canvas texture
Tap a tag to carry it over to the builder
Negative Prompt
digital art, 3d, photo, anime, modern, cartoon, smooth gradient, airbrush, vector
Tag Breakdown — Why these tokens?
Three-layer medium lock — differentiates from digital painting
Historic lighting technique — triangular cheek highlight
Period + genre compressed — summons the composition vocabulary
Wardrobe + prop give the scene a narrative hook
Palette lock — classical painting’s signature earthen tone
Variations
- ▸ Landscape: `renaissance portrait` → `landscape painting, english countryside, turner style`
- ▸ Impressionism: `rembrandt lighting` → `impressionist, monet style, soft brush dabs`
- ▸ Still life: `1 woman, renaissance portrait` → `still life, fruit bowl, wine bottle, 17th century dutch style`
- ▸ Contemporary: `classical oil painting` → `contemporary oil painting, loose gestural strokes`
Recommended Model
SDXL Base / DreamShaper XL
SDXL reproduces oil texture natively without extra LoRAs
Related tag categories
Dark Fantasy Character
A dark fantasy character inspired by the Soulsborne or Berserk universe. Medieval armor, broken weapons, and scenes dusted with blood and ash. Keep the background intentionally murky and the mood heavy. An earth-tone plus blood-red palette conveys heroism and tragedy at once.
Prompt
masterpiece, best_quality, dark fantasy, 1 knight, solo, full body, battered plate armor, cracked helmet, broken greatsword, blood splatter, ash-covered ruins, foggy battlefield, dim moonlight, muted earth tones, crimson accents, dramatic pose, berserk inspired, volumetric fog, cinematic
Tap a tag to carry it over to the builder
Negative Prompt
bright colors, cheerful, cute, chibi, modern clothing, sci-fi, futuristic, low quality, anime chibi
Tag Breakdown — Why these tokens?
Genre + reference immediately locks the tone
Three battered props — the narrative that the subject has fought is visible
Environmental trauma cues — hint at the character’s past
Palette contrast — earth tone + blood red
Atmosphere + lighting deliver depth and dread simultaneously
Variations
- ▸ Witch variant: `1 knight, battered plate armor` → `1 witch, tattered robe, dark staff, glowing sigils`
- ▸ Monster hunter: `broken greatsword` → `silver weapon, monster parts, witcher inspired`
- ▸ Heroic beat: `dramatic pose` → `kneeling, helmet removed, single tear`
- ▸ Castle setting: `ash-covered ruins` → `cursed castle interior, chandelier, dusty thrones`
Recommended Model
SDXL + dark fantasy LoRA / Juggernaut XL
Grain and dark tones stabilize best on XL-class checkpoints
Related tag categories
Slice-of-Life Scene
An ordinary afternoon, a small pocket of happiness — the slice-of-life aesthetic. Without dramatic composition, spatial detail and soft lighting build a sense of lived-in warmth. Think Makoto Shinkai summer afternoons; emotional tags like `cozy`, `soft light`, and `ordinary day` do the heavy lifting.
Prompt
masterpiece, best_quality, slice of life, 1girl, solo, sitting by a window, reading a book, morning light, coffee cup, cozy apartment, plants on windowsill, short brown hair, oversized sweater, peaceful expression, soft color palette, makoto shinkai inspired, gentle atmosphere, ordinary day
Tap a tag to carry it over to the builder
Negative Prompt
dramatic, action, epic, battle, dark, scary, low quality, cluttered, oversaturated
Tag Breakdown — Why these tokens?
Genre + reference lock the emotional tone
Two actions build a micro-narrative — "she is reading"
Light and plants — signatures of a lived-in space
Highlights the subject’s casual ease
Double-specifies the spatial emotion for stability
Variations
- ▸ Cafe: `cozy apartment` → `small cafe, pastries on table, rainy window`
- ▸ Kitchen: `reading a book` → `cooking, steam from pot, kitchen apron`
- ▸ Pet add-on: append `a cat sleeping on her lap`
- ▸ Autumn: `morning light` → `autumn afternoon, warm golden light`
Recommended Model
Animagine XL / Counterfeit / Pastel Mix
Anime checkpoints’ soft palettes suit slice-of-life best
Related tag categories
Pixel Art
A pixel-art prompt channeling retro game aesthetics. Explicitly constraining resolution and palette (`pixel art`, `16-bit`, `limited palette`) nudges the model toward genuine pixel expression. Perfect single-pixel alignment is still hard — the production workflow is to finish in Aseprite or Pixaki.
Prompt
masterpiece, pixel art, 16-bit, retro game style, 1 hero character, solo, sword and shield, medieval fantasy village, cobblestone street, torches, small houses, limited palette, isometric view, SNES era aesthetic, detailed sprite work, clear outlines
Tap a tag to carry it over to the builder
Negative Prompt
realistic, photo, 3d, smooth shading, blurry, anti-aliased, modern graphics, high resolution blur
Tag Breakdown — Why these tokens?
Medium + era + palette — three-axis style lock
Hardware reference fixes tone — mid-90s visual vocabulary
Viewpoint lock — the RPG canonical angle
Pixel quality signal — suppresses blurry outputs
Three elements that complete the setting
Variations
- ▸ Sci-fi pixel: `medieval fantasy village` → `cyberpunk space station, neon lights, 16-bit sci-fi`
- ▸ 8-bit: `16-bit, SNES era` → `8-bit, NES era, 4-color palette`
- ▸ Battle scene: `1 hero character` → `2 fighters, battle stance, particle effects`
- ▸ Interior: `cobblestone street` → `pixel art tavern interior, fireplace, wooden beams`
Recommended Model
SDXL + pixel art LoRA / Pixel Art Diffusion
A dedicated pixel LoRA is essential for pixel accuracy
Related tag categories
3D Render Portrait
A 3D character portrait with the polish of Unreal Engine or Octane. Covers the stylized end of 3D through to hyperreal CG, and CG-industry vocabulary like `subsurface scattering`, `ray tracing`, `PBR materials` raises render quality. Ideal for character asset sheets or game trailer stills.
Prompt
masterpiece, best_quality, 8k, 3d render, stylized 3d animation, 1 boy, solo, teenage, curly red hair, freckles, green eyes, adventurer outfit, leather jacket, confident smile, looking at viewer, soft studio lighting, subsurface scattering, ray tracing, PBR materials, Unreal Engine 5, Octane render, character sheet
Tap a tag to carry it over to the builder
Negative Prompt
2d, anime, painting, flat shading, low poly, old graphics, blurry, photo realistic (unless intended)
Tag Breakdown — Why these tokens?
Medium + render treatment locks the style
Three high-end CG rendering quality cues
Two renderers specified — pushes for maximum detail
Three identifiable traits — what an asset sheet needs
Usage cue — enforces a clean, on-model portrait angle
Variations
- ▸ Photoreal 3D: `stylized 3d animation` → `hyperrealistic 3d scan, cinema 4d, mental ray`
- ▸ Full body: `character sheet` → `full body turnaround, T-pose, asset reference`
- ▸ With scene: append `fantasy tavern background, prop set dressing`
- ▸ Older subject: `teenage, curly red hair` → `elderly man, grey beard, weathered skin`
Recommended Model
SDXL Base / RealVisXL / 3D styles LoRA
Prompting alone works; a LoRA improves consistency across sessions
Related tag categories
Nature Photography
Landscape photography aiming for nature-documentary stills. Combine location + time + natural phenomenon for a unique moment, and declaring camera specs (Nikon, 24-70mm, long exposure) secures photographic authenticity. For wildlife, swap to telephoto setups like `wildlife photography, telephoto 400mm`.
Prompt
masterpiece, best_quality, 8k, photorealistic, landscape photography, autumn forest, golden maple leaves, misty river, morning fog, sun rays piercing through trees, long exposure, Nikon D850, 24-70mm lens, f/8, vibrant yet natural colors, high dynamic range, documentary nature photography
Tap a tag to carry it over to the builder
Negative Prompt
people, buildings, cars, cartoon, anime, painting, oversaturated, blurry, low detail, watermark
Tag Breakdown — Why these tokens?
Genre + shooting style locks the photographic vocabulary
Three-layer combo: place + season + natural element
Lighting phenomenon — the dramatic natural-light moment
Camera-spec cues — authenticity for a photographic look
Tone-mapping direction: rich but not over-the-top
Variations
- ▸ Arctic: `autumn forest, golden maple` → `arctic landscape, aurora borealis, snow plains`
- ▸ Beach: `misty river, morning fog` → `tropical beach, crashing waves, sunset`
- ▸ Wildlife: `landscape photography` → `wildlife photography, eagle in flight, telephoto 400mm`
- ▸ Astro: `sun rays` → `milky way, starry sky, astrophotography, 30 second exposure`
Recommended Model
SDXL Realistic / RealVisXL / Juggernaut XL
Landscape photorealism requires a realistic checkpoint
Related tag categories
Build your own variation
Every prompt above was assembled from Promgrammer's curated tag library. Open the builder, pick any example as a starting point, then swap in tags from the 18 categories to converge on your vision. Save your best prompts for reuse, favorite the tags that consistently work, and let the copy history track what you've tried.