Prompt Library

AI Image Prompts That Actually Produce Professional Results

40 copy-paste prompts

40 tested prompts for Midjourney, DALL-E, and Stable Diffusion that go beyond "a cat in space." Learn the prompt engineering techniques that separate amateur AI images from portfolio-worthy visuals.

In short: This page contains 40 copy-paste ready prompts, organized into 8 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

Photography Style

5 prompts

Cinematic Portrait with Natural Light

1/40

✨ What it does

This prompt builds a Midjourney-ready cinematic portrait of a [age]-year-old [ethnicity] [man/woman] in golden-hour window light, shot like Kodak Portra. You fill [describe expression], generate a few frames, and pick the one you would actually use.

A cinematic portrait of a [age]-year-old [ethnicity] [man/woman] photographed with a Canon EOS R5, 85mm f/1.4 lens, golden hour natural light streaming through a window. Shallow depth of field, film grain, warm color grading reminiscent of Kodak Portra 400. The subject has [describe expression] and is wearing [describe clothing]. Shot composition follows the rule of thirds with the subject positioned at the left intersection point.

💡

Pro tip: Naming specific camera bodies and lenses triggers the AI to mimic the optical characteristics of that equipment — bokeh shape, lens compression, and color rendering all change based on what you specify.

Dramatic Landscape with Atmosphere

2/40

✨ What it does

This prompt builds a wide landscape of [location/terrain type] at [time of day], with fog, dramatic clouds, and a long exposure on [water/clouds]. You choose [warm/cool] light, generate a few versions, and keep the moodiest one.

A breathtaking landscape photograph of [location/terrain type] at [time of day]. Volumetric fog rolling through a valley, dramatic cloud formations lit from below by [warm/cool] light. Shot with a Sony A7R IV, 16-35mm f/2.8 wide-angle lens at f/11 for maximum sharpness. Long exposure effect on [water/clouds]. The color palette is dominated by [primary color] and [secondary color]. National Geographic quality, award-winning nature photography.

💡

Pro tip: Adding "National Geographic quality" or "award-winning" as style references consistently produces higher-quality compositions. Specifying f-stop values controls the depth of field in the generated image.

Street Photography with Story

3/40

✨ What it does

This prompt builds a rainy-night street photo in [city name]: wet cobblestones, neon reflections, a lone figure with an umbrella, shot like a Leica 35mm. You set the mood, generate a few frames, and pick the one that feels like a story.

A candid street photograph taken in [city name] on a rainy evening. Wet cobblestone streets reflecting neon signs and streetlights. A lone figure with an umbrella walks away from the camera. Shot on a Leica M10 with a 35mm Summicron lens, ISO 1600, slight motion blur suggesting movement. The mood is [melancholic/mysterious/nostalgic]. Black and white with selective color on the [red umbrella/yellow taxi/blue neon sign]. Henri Cartier-Bresson inspired decisive moment.

💡

Pro tip: Referencing specific photographers by name guides the AI toward their compositional style. Cartier-Bresson gives you geometric composition, Saul Leiter gives you color and abstraction, Vivian Maier gives you intimate street scenes.

Macro Nature Close-Up

4/40

✨ What it does

This prompt builds an extreme macro of [a dewdrop on a spider web / a butterfly wing / a flower petal], razor-sharp on the subject with creamy [color] blur behind. You pick the subject, generate a few close-ups, and keep the sharpest one.

An extreme macro photograph of [a dewdrop on a spider web / a butterfly wing / a flower petal]. Shot with a Canon MP-E 65mm macro lens at 5:1 magnification. Crystal-clear focus on the subject with a completely blurred creamy bokeh background in [color]. Water droplets act as tiny lenses reflecting the surrounding [garden/forest]. Studio-quality ring light illumination revealing iridescent textures and microscopic details invisible to the naked eye. 8K resolution, focus stacking technique.

💡

Pro tip: Mentioning "focus stacking" tells the AI to render the subject with front-to-back sharpness rather than the paper-thin depth of field typical in macro shots. This produces much more detailed results.

Environmental Portrait in Context

5/40

✨ What it does

This prompt builds an environmental portrait of someone at work in their space, caught mid-action in available light. You fill the job, the action, and the light source, then generate until the room tells the story.

An environmental portrait of a [profession — e.g., blacksmith, chef, fisherman, potter] in their workspace. The subject is captured mid-action, [describe the specific action]. Shot with available light from [describe light source — e.g., forge fire, kitchen window, harbor dawn]. Fujifilm X-T5 with a 56mm f/1.2 lens, slightly desaturated color grading with lifted blacks. The workspace fills the frame with authentic tools and materials, telling the story of the craft. Documentary photography style, Steve McCurry inspired composition.

💡

Pro tip: Environmental portraits work best when you describe the light source as part of the scene rather than as a photography term. "Forge fire lighting the face" is more effective than "warm key light from camera left."

Illustration & Digital Art

5 prompts

Modern Flat Design Illustration

6/40

✨ What it does

This prompt builds a flat vector scene with [color 1], [color 2], [color 3], and [color 4], no gradients, and simple rounded people. You describe the scene, generate a few layouts, and pick the cleanest one for a slide or site.

A flat design vector illustration of [scene/concept — e.g., a team collaborating in a modern office]. Clean geometric shapes, limited color palette of [color 1], [color 2], [color 3], and [color 4] with white space. No gradients, no shadows, no outlines. Characters have simple rounded features with no facial details beyond basic shapes. Isometric perspective. The composition is balanced and suitable for a [website hero section / app onboarding screen / blog header]. Dribbble trending style, 2026 design aesthetic.

💡

Pro tip: Specifying "no gradients, no shadows, no outlines" is essential for achieving true flat design. Without these constraints, AI generators tend to add subtle dimensionality that breaks the flat aesthetic.

Editorial Magazine Illustration

7/40

✨ What it does

This prompt builds a magazine illustration that talks about a topic through a visual metaphor, not a literal picture. You name the topic and the metaphor, generate a few concepts, and pick the one that would sit next to an article.

An editorial illustration for a magazine article about [topic — e.g., the future of remote work, climate change, digital privacy]. Conceptual and metaphorical — do not illustrate the topic literally. Instead, use visual metaphor: [describe the metaphor — e.g., a person building a bridge between two floating islands]. Textured brushwork with visible strokes, muted sophisticated color palette, slight paper texture overlay. Style inspired by [New Yorker covers / Monocle magazine / The Economist]. Horizontal composition with space for text overlay on the [left/right] third.

💡

Pro tip: The key to great editorial illustration prompts is telling the AI what NOT to do. "Do not illustrate literally" forces the AI to think conceptually, which produces far more interesting and publishable results.

Children's Book Illustration

8/40

✨ What it does

This prompt builds a gentle children's-book scene in soft watercolor and pencil, with a warm [dominant color]. You describe the scene, generate a few pages, and pick the one that feels kind enough for a kid.

A whimsical children's book illustration of [scene — e.g., a small fox discovering a hidden library inside an ancient oak tree]. Soft watercolor textures with pencil line work underneath. Warm, inviting color palette with [dominant color] tones. The style is gentle and detailed, inspired by [Beatrix Potter / Studio Ghibli / Oliver Jeffers]. The scene includes many small discoverable details that reward close looking — [describe 2-3 hidden details]. The lighting is [soft morning light / warm lamplight / dappled forest light]. Suitable for ages 4-8.

💡

Pro tip: Adding "many small discoverable details" is a powerful technique that makes AI-generated illustrations much more engaging. It triggers the model to add the kind of rich environmental storytelling that defines great children's book art.

Retro Vintage Poster Design

9/40

✨ What it does

This prompt builds a 1930s Art Deco travel poster for [real or fictional destination], with a short color set and the name [Name] locked at the [top/bottom]. You fill those, generate a few posters, and pick the one that still reads at a glance.

A vintage travel poster for [real or fictional destination] in the style of 1930s Art Deco tourism posters. Bold geometric shapes, limited screen-print color palette of 4-5 colors with [dominant color]. Strong vertical composition with the destination name "[Name]" integrated into the design at the [top/bottom]. Stylized [landmark or landscape feature] as the focal point. Slight paper texture and print registration marks for authenticity. Inspired by the work of A.M. Cassandre and the WPA poster series.

💡

Pro tip: For vintage styles, always specify the number of colors and mention print technique limitations like "screen-print" or "lithograph." This constrains the AI to produce results that look like they could have actually been printed with period-accurate methods.

Infographic Icon Set

10/40

✨ What it does

This prompt builds a 3x3 set of matching line icons about your topic, all the same stroke weight, in [single accent color] on white. You list the nine ideas, generate the grid, and use the set together so the styles match.

A cohesive set of 9 icons arranged in a 3x3 grid for an infographic about [topic — e.g., sustainable energy, digital marketing, health and wellness]. Each icon represents: [list 9 concepts]. Style: line icons with consistent 3px stroke weight, rounded caps, [single accent color] on white background. Each icon fits within a 64x64px conceptual boundary. The set should feel unified — same level of detail, same visual weight, same degree of abstraction. Modern, minimal, Silicon Valley startup aesthetic.

💡

Pro tip: Specifying exact stroke weight and grid size helps the AI produce icons that actually look like they belong together. Without these constraints, each icon tends to have a slightly different level of detail and weight.

Concept Art & Fantasy

5 prompts

Epic Fantasy Environment

11/40

✨ What it does

This prompt builds a huge fantasy landscape with tiny figures for scale and dramatic light through fog or glow. You describe the place, generate a few wide shots, and pick the one that feels big enough for a game or film board.

A sweeping concept art painting of [describe environment — e.g., a floating city connected by crystal bridges above an endless cloud sea]. Epic scale with tiny human figures for size reference. Dramatic volumetric lighting with [god rays / bioluminescence / aurora borealis] illuminating the scene. Rich atmospheric perspective with foreground, midground, and background layers. Color palette shifts from [warm colors] in the foreground to [cool colors] in the distance. Painted in the style of a AAA video game environment — think Elden Ring meets Studio Ghibli. Matte painting quality, 16:9 cinematic aspect ratio.

💡

Pro tip: Including "tiny human figures for size reference" is a concept art technique that dramatically improves the sense of scale. Without it, the AI has no anchor and environments often feel miniature rather than epic.

Character Design Sheet

12/40

✨ What it does

This prompt builds a character sheet with front, three-quarter, and back views, plus callouts for a weapon, a clothing detail, and an accessory. You fill the character type and look, generate the sheet, and use it as a drawing reference.

A character design sheet for a [character type — e.g., rogue alchemist, cyberpunk street samurai, elven diplomat]. Show the character in three views: front, three-quarter, and back. Include detail callouts for [weapon/tool], [distinctive clothing element], and [accessory]. The character has [physical description]. Their color palette is [3-4 colors]. Art style: [semi-realistic / anime-influenced / western comic]. The design should tell a story about who this character is through visual details — worn edges on equipment, cultural motifs in clothing, scars or modifications that hint at backstory. White background, clean presentation.

💡

Pro tip: The phrase "tell a story through visual details" is crucial for character design. It pushes the AI beyond generic fantasy archetypes and into designs with personality and implied narrative.

Creature Design with Anatomy

13/40

✨ What it does

This prompt builds a predator for a habitat by mixing [real animal 1] and [real animal 2] in a way that could almost exist, shown mid-hunt. You pick the two animals, generate a few creatures, and keep the one that still looks alive.

A creature design for a [habitat — e.g., deep ocean, volcanic cave, frozen tundra] predator. The creature combines anatomical features of [real animal 1] and [real animal 2] in a biologically plausible way. Show the creature in a dynamic pose that demonstrates its primary hunting behavior. Bioluminescent [markings/eyes/tendrils] in [color] contrast against [dark/muted] body coloring. Include a human silhouette scale comparison in the corner. Concept art style with loose painterly rendering, visible brushstrokes. The design should feel like it could exist in nature — strange but not impossible.

💡

Pro tip: Grounding creature design in real animal anatomy is the secret to believable monsters. "Combines features of a mantis shrimp and a snow leopard" produces far more interesting results than "scary monster with claws and teeth."

Sci-Fi Vehicle or Mech Design

14/40

✨ What it does

This prompt builds a detailed sci-fi vehicle or mech at a three-quarter angle, with a person for scale. You choose the vehicle type and whether it looks military, sleek, or scavenged, then generate until the parts look like they work.

A detailed concept art illustration of a [vehicle type — e.g., single-pilot reconnaissance mech, deep-space mining vessel, hover bike]. The design philosophy is [utilitarian military / sleek civilian / cobbled-together scavenger]. Show the vehicle at a three-quarter angle with [pilot/driver] visible for scale. Key design features: [list 3 distinctive features — e.g., asymmetric weapon mount, exposed reactor core, modular cargo pods]. Surface details include [weathering/panel lines/decals/rust]. Environment context: parked in [setting]. Lighting: [dramatic hangar lighting / harsh desert sun / neon-lit rain]. Syd Mead meets Ian McQue aesthetic.

💡

Pro tip: Specifying a "design philosophy" like "utilitarian military" vs. "cobbled-together scavenger" completely changes the visual language. Military designs have clean lines and standardized parts. Scavenger designs have mismatched components and improvised repairs. This single word choice transforms the entire output.

Fantasy Map Illustration

15/40

✨ What it does

This prompt builds a hand-drawn parchment map of [continent/region name] in a Tolkien-like style, with mountains, forests, rivers, and named places. You list the locations, generate a few maps, and pick the one you would hang in a book.

A hand-drawn fantasy map of [continent/region name] in the style of classic Tolkien-era cartography. Parchment background with aged edges and coffee stain effects. Mountain ranges drawn in profile view, forests represented by clusters of tiny trees, rivers flowing from mountains to coast. Include: [list 5-7 named locations with types — e.g., "Thornhaven (fortified city)", "The Whispering Marshes (swamp)", "Mount Ashcrown (volcano)"]. A decorative compass rose in the [corner]. The map title "[Name]" in ornate calligraphic lettering with a decorative cartouche border. Sea areas filled with subtle wave patterns and a [sea monster/ship] illustration.

💡

Pro tip: Listing specific named locations with their types gives the AI enough worldbuilding context to place them logically — ports on coastlines, fortresses at mountain passes, trading cities at river crossings. The map tells a story about the world through geography.

Like these prompts? There are full tutorials behind them.

Learn the workflows, not just the prompts. 300+ easy-to-follow tutorials inside AI Academy — and growing every week.

Try AI Academy Free

Product & Marketing

5 prompts

Premium Product Hero Shot

16/40

✨ What it does

This prompt builds a studio hero shot of a product on a chosen surface, with a key light and a clean commercial look. You name the product and surface, generate a few frames, and keep the one that could sit on a product page.

A premium product photography shot of a [product — e.g., matte black wireless headphones, artisan coffee bag, luxury skincare bottle] on a [surface material — e.g., polished marble, raw concrete, dark wood]. Studio lighting setup: key light at 45 degrees creating a [dramatic shadow / soft gradient], fill light from the opposite side at 30% intensity, and a hair light creating a rim highlight that separates the product from the background. Background is [solid color / gradient]. The product is the hero — no distracting elements. Shot would be used for [e-commerce listing / brand website hero / Instagram ad]. Apple-level product photography quality.

💡

Pro tip: Describing the lighting setup in photography terms (key light, fill light, hair light) produces dramatically better results than just saying "well lit." The AI understands studio lighting terminology and applies it accurately.

Lifestyle Product in Context

17/40

✨ What it does

This prompt builds a lifestyle photo where [product] is used naturally in a real room, not isolated on white. You describe the person and the setting, generate a few moments, and pick the one that still looks lived-in.

A lifestyle photograph showing [product] being used naturally by a [describe person — age, style, energy] in a [setting — e.g., sun-filled Scandinavian apartment, bustling outdoor cafe, modern home gym]. The product is visible but not the sole focus — it is integrated into a desirable lifestyle moment. The person is [describe natural action — e.g., laughing while pouring coffee, mid-stretch with earbuds in]. Warm, aspirational color grading. Shot on a 50mm lens with a natural depth of field that keeps the person and product sharp while the background is softly blurred. The image should make the viewer think "I want that life" not just "I want that product." Everlane / Aesop brand photography aesthetic.

💡

Pro tip: Naming specific brands whose photography style you admire is more effective than describing the aesthetic in abstract terms. "Everlane aesthetic" communicates more about lighting, composition, and mood than a paragraph of description.

Social Media Carousel Slide

18/40

✨ What it does

This prompt builds a square social slide with [Key statistic or headline text] in bold type on a [brand color] field with [contrasting color] text. You generate a few layouts and pick the one that stays readable at a glance.

A clean, modern social media carousel slide (1080x1080px aspect ratio) for [platform — Instagram/LinkedIn]. The slide presents: "[Key statistic or headline text]" in bold [sans-serif font style] text. Color scheme: [brand color] background with [contrasting color] text. Include a subtle [geometric pattern / gradient mesh / abstract shape] as a background design element. The text hierarchy is clear: large headline number/word, smaller supporting text beneath. White space is generous — the slide should breathe. No photographs, no illustrations — pure typographic design. The style matches [brand reference — e.g., HubSpot, Notion, Linear] social content.

💡

Pro tip: AI image generators handle typographic designs better when you specify "no photographs, no illustrations" to prevent the model from adding unwanted visual elements. Keep the design concept simple and let negative space do the work.

Food Photography Flat Lay

19/40

✨ What it does

This prompt builds a top-down food photo of a dish on a chosen surface, with a few props and a slightly imperfect, styled look. You name the meal and the table, generate a few layouts, and keep the one you would post.

A top-down flat lay food photograph of [dish/meal — e.g., artisan sourdough toast with avocado and poached eggs, a holiday cookie decorating spread, a charcuterie board]. Arranged on a [surface — e.g., weathered wood table, white marble, linen tablecloth]. Surrounded by complementary props: [list 3-4 items — e.g., scattered herbs, a linen napkin, vintage silverware, a cup of black coffee]. Natural window light from the [top-left/top-right] creating soft shadows. The color story is [warm earth tones / bright and fresh / moody and dark]. Styled with intentional imperfection — a few crumbs, a slightly askew napkin, a bite taken from one piece. Bon Appetit magazine quality food styling.

💡

Pro tip: The phrase "intentional imperfection" is the key to believable food photography. Without it, AI produces impossibly perfect arrangements that look artificial. A few crumbs and an askew element add the authenticity that makes food images feel inviting.

App UI Mockup in Device

20/40

✨ What it does

This prompt builds a photo of a phone in a hand showing your app screen, slightly angled, so it looks like a real device shot. You describe the screen, generate a few mockups, and pick the one that does not look pasted on.

A photorealistic mockup of a mobile app displayed on an [iPhone 16 Pro / Samsung Galaxy S26] held naturally in a person's hand. The screen shows [describe the app screen — e.g., a meditation app with a calming gradient, timer display, and minimal UI]. The device is at a slight angle, not perfectly flat. The background is [blurred office / cafe / outdoor setting] with bokeh. Lighting is natural and the screen brightness looks realistic — not an obviously pasted screenshot. The hand holding the phone has [natural skin tone], and the grip looks comfortable and authentic. Product launch marketing quality.

💡

Pro tip: Specifying that the device should be "at a slight angle, not perfectly flat" and that "the grip looks comfortable" prevents the AI from producing the stiff, unnatural device-in-hand images that immediately look like stock photography.

Architecture & Interior

5 prompts

Modern Architectural Visualization

21/40

✨ What it does

This prompt builds a photoreal render of a building in a real landscape, with people-scale details so it feels lived in. You name the building type and the setting, generate a few views, and keep the one that feels like a place.

An architectural visualization render of a [building type — e.g., minimalist concrete villa, glass-walled museum, timber-frame mountain cabin] set in [environment — e.g., a coastal cliff, a dense forest, an urban lot]. The building features [key architectural elements — e.g., cantilevered upper floor, floor-to-ceiling glazing, green roof with native plantings]. Golden hour lighting with long shadows emphasizing the building's form. Surrounding landscape is [describe landscaping]. Include subtle signs of life — [a bicycle leaning against the wall, warm interior lights visible, an open book on the terrace]. Architectural Digest quality, V-Ray render aesthetic. Shot from a [low angle emphasizing height / eye-level perspective / aerial three-quarter view].

💡

Pro tip: The "subtle signs of life" instruction is what separates compelling architectural visualization from cold CAD renders. A warm light in a window or a plant on a terrace tells the viewer this is a place for living, not just a geometric exercise.

Interior Design Mood Shot

22/40

✨ What it does

This prompt builds a magazine interior of a room in a named design style, with a few hero furniture pieces. You pick the room and style, generate a few angles, and save the one that matches the mood you want.

A stunning interior photograph of a [room type — e.g., living room, master bedroom, chef's kitchen] designed in [style — e.g., Japanese minimalism, mid-century modern, Scandinavian hygge, industrial loft]. Key furniture pieces: [list 2-3 hero items — e.g., a Noguchi coffee table, a modular linen sectional, a statement pendant light]. Material palette: [list materials — e.g., white oak floors, plaster walls, brushed brass fixtures, natural stone]. The room is flooded with [natural light from large windows / warm ambient evening light / a mix of natural and designed lighting]. Color palette centers on [describe 2-3 colors]. Include [living elements — a thriving monstera plant, a stack of design books, a casually draped throw blanket]. The shot is styled for Architectural Digest — aspirational but not impossibly perfect.

💡

Pro tip: Naming real furniture pieces (Noguchi table, Eames chair) grounds the AI in actual design language and produces results that look like real interiors rather than generic renders. The AI knows what these pieces look like and places them appropriately.

Before/After Renovation Concept

23/40

✨ What it does

This prompt builds a side-by-side before-and-after of the same room from the same camera angle. You describe the current mess and the finished look, generate the pair, and use it in a pitch or a portfolio.

A side-by-side before and after renovation concept for a [room/space — e.g., cramped galley kitchen, dated 1990s bathroom, dark basement]. LEFT side (before): [describe current state — e.g., dark wood cabinets, laminate counters, fluorescent lighting, cluttered countertops, vinyl flooring]. RIGHT side (after): the same room transformed with [describe renovation — e.g., open shelving, quartz waterfall countertops, recessed LED lighting, minimalist decor, herringbone tile floors]. Both views from the exact same camera angle to emphasize the transformation. The "after" side should feel realistic and achievable, not fantasy — a renovation you could actually do, not a complete structural rebuild. Consistent lighting between both sides.

💡

Pro tip: Emphasizing "realistic and achievable" prevents the AI from showing a complete structural transformation. The most compelling before/afters keep the same bones and show how finishes, lighting, and styling completely transform a space without moving walls.

Outdoor Living Space Design

24/40

✨ What it does

This prompt builds an evening outdoor living space with seating, light, and a place you would actually sit. You fill the setting and the main features, generate a few views, and pick the one that feels like a night outside.

A beautiful outdoor living space design for a [setting — e.g., rooftop terrace in a city, backyard patio with pool, covered veranda overlooking mountains]. The space features [key elements — e.g., a built-in outdoor kitchen with stone counters, a fire pit conversation area, a pergola with climbing wisteria]. Furniture: [describe outdoor furniture — e.g., teak dining set, modular outdoor sofa with all-weather cushions, hanging egg chairs]. The lighting plan includes [string lights / lanterns / recessed step lighting / landscape uplighting on trees] creating a warm evening ambiance. Season: [describe — e.g., late summer evening, autumn afternoon]. Include [atmospheric details — e.g., a pitcher of sangria on the table, throw blankets draped on chairs, candles flickering]. Shot at dusk when both natural and artificial light create the most magical atmosphere.

💡

Pro tip: Specifying "shot at dusk" for outdoor spaces produces the most appealing results because the AI renders both the remaining natural light and the artificial lighting, creating the layered warmth that makes outdoor spaces feel magical in real life.

Boutique Hotel Room Concept

25/40

✨ What it does

This prompt builds a boutique hotel room that feels tied to a place, with local materials and one signature feature. You name the culture and the signature, generate a few rooms, and keep the one that does not look like a generic suite.

An interior design concept for a boutique hotel room inspired by [location/culture — e.g., coastal Greek island, Tokyo minimalism, Marrakech riad, Tulum jungle]. The room features [signature element — e.g., an arched doorway to a private terrace, a sunken bathtub with a view, a canopy bed with sheer linen drapes]. Local materials and craftsmanship: [describe — e.g., hand-troweled tadelakt walls, locally woven textiles, carved wood screen panels]. The color palette draws from the local environment: [describe 3 colors and their inspiration]. Morning light enters through [describe windows/openings]. The room balances luxury with authenticity — it feels like a place, not a hotel chain. Include one carefully chosen [artwork / artifact / plant] that connects the room to its location. Cereal Magazine travel photography aesthetic.

💡

Pro tip: The instruction to "balance luxury with authenticity" is critical for boutique hotel design. Without it, the AI defaults to either generic luxury hotel aesthetic or an over-the-top themed room. The best boutique hotels feel like elevated versions of local living, not stage sets.

Go from copy-pasting to actually mastering AI.

AI Academy: 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ other tools. New tutorials added every week.

Start Your Free Trial

Abstract & Artistic

5 prompts

Large-Scale Abstract Painting

26/40

✨ What it does

This prompt builds a large abstract painting with thick paint, thin washes, and a chosen color set. You set the colors and the feeling, generate a few canvases, and pick the one you would print or hang.

A large-scale abstract expressionist painting dominated by [color palette — e.g., deep navy, burnt sienna, and metallic gold]. The composition features bold gestural brushstrokes with visible texture — thick impasto areas contrasting with thin washes where the canvas shows through. The painting evokes [emotion/concept — e.g., the tension between chaos and order, the feeling of standing at the edge of the ocean]. No recognizable objects or figures. The work references [artist influence — e.g., the scale of Rothko, the energy of de Kooning, the color field approach of Helen Frankenthaler]. Gallery-quality artwork photographed on a white wall with museum lighting, slight shadow beneath the canvas suggesting physical depth. Aspect ratio 3:4, vertical orientation.

💡

Pro tip: Abstract art prompts work best when you describe the emotional intent alongside the technical approach. "Evokes the tension between chaos and order" gives the AI a compositional direction that "random abstract painting" never will.

Generative Art Pattern

27/40

✨ What it does

This prompt builds a generative pattern from a simple shape that grows by a math rule, shifting as it moves across the frame. You pick the shape and the growth idea, generate a few fields, and keep the one that still looks designed.

A generative art composition created from [geometric primitive — e.g., circles, triangles, hexagons] following a mathematical growth pattern inspired by [phyllotaxis / Voronoi diagrams / fractal branching / Perlin noise flow fields]. The elements transition in [size/color/opacity/rotation] from [one edge/the center] to [the opposite edge/the perimeter]. Color palette: [describe gradient or palette — e.g., a gradient from electric blue to coral pink through violet]. Background: [pure black / deep navy / white]. The overall impression is organic despite being mathematically generated — like discovering geometry in nature. High contrast, crisp edges, suitable for large-format printing at 300 DPI. Inspired by the work of [Tyler Hobbs / Manolo Gamboa Naon / Vera Molnar].

💡

Pro tip: Referencing specific generative artists produces dramatically different results. Tyler Hobbs gives you flowing, organic compositions. Vera Molnar gives you systematic grid-based variations. Manolo Gamboa Naon gives you dense, colorful complexity. The artist reference is effectively a style parameter.

Surrealist Photomanipulation

28/40

✨ What it does

This prompt builds a photoreal picture of an impossible scene, sharp like a photo and wrong like a dream. You describe the impossibility, generate a few versions, and pick the one that still looks physically real.

A surrealist photomanipulation where [describe the impossible scene — e.g., a staircase spirals upward out of a teacup into clouds, a person's silhouette is filled with a forest landscape, gravity is reversed in a library with books floating upward]. The image should be photorealistic in its rendering but impossible in its physics — the surrealism comes from the concept, not from distorted rendering. Seamless compositing with matched lighting and shadows across all elements. Muted, slightly desaturated color palette with one [vibrant accent color] drawing the eye to the focal point. The mood is [dreamlike / unsettling / whimsical / contemplative]. Inspired by [Erik Johansson / Rob Gonsalves / Rene Magritte]. Suitable for gallery print or album cover.

💡

Pro tip: Specifying "photorealistic rendering but impossible physics" is the key instruction that separates sophisticated surrealism from random AI weirdness. The power of surrealist art is the contrast between how real it looks and how impossible it is.

Mixed Media Texture Collage

29/40

✨ What it does

This prompt builds a mixed-media collage from torn paper, paint, ink, and other handmade layers around a theme. You list the materials and the theme, generate a few pieces, and keep the one that looks touched by a hand.

A mixed media art piece combining [list 3-4 media — e.g., torn vintage newspaper clippings, watercolor washes, ink line drawings, gold leaf accents, pressed flowers, fabric swatches]. The composition is loosely centered on the theme of [theme — e.g., memory, urban decay, the passage of time, botanical study]. Layered textures with visible edges where different media overlap. Some elements are precisely rendered while others are loose and gestural. The background is [describe — e.g., aged kraft paper, gesso-textured canvas, old book pages]. Neutral base palette with pops of [accent color]. The piece should look like it was created by hand — imperfect registration, visible glue edges, slightly wrinkled paper. Wabi-sabi aesthetic, the beauty of imperfection.

💡

Pro tip: The instruction "should look like it was created by hand" combined with specific imperfections (wrinkled paper, visible glue edges) is what makes AI-generated mixed media look authentic. Without it, the AI produces impossibly clean collages that look obviously digital.

Light Painting Long Exposure

30/40

✨ What it does

This prompt builds a long-exposure light-painting photo with glowing trails against a night background. You choose the lights, the pattern, and [word] if you want text, then generate until the trails feel drawn.

A long-exposure light painting photograph against a [pitch black / twilight / urban night] background. Trails of [colored LED lights / steel wool sparks / sparklers] create [geometric patterns — perfect circles, spirals, figure eights / organic flowing forms / written text spelling "[word]"]. Multiple exposures layered to create depth and complexity. The light trails are [sharp and precise / soft and ethereal / a mix of both]. Color palette of the light: [describe — e.g., warm amber and cool cyan creating complementary contrast]. If a human figure is present, they appear as a [ghostly motion blur / sharp silhouette / partially visible form]. The image captures something invisible to the naked eye — time itself made visible through accumulated light. DSLR long-exposure quality, no post-processing look.

💡

Pro tip: Light painting prompts benefit from describing both what the light does AND what the darkness does. "Pitch black background" gives the light trails maximum contrast and visual impact. Adding a twilight background creates a completely different, more atmospheric mood.

Prompt Engineering Techniques

5 prompts

Build a Detailed Prompt from a Concept

31/40

✨ What it does

ChatGPT expands your one-sentence idea into a full image prompt for [Midjourney / DALL-E / Stable Diffusion], with subject, light, camera, and style spelled out. You paste that prompt into the generator and tweak whatever still looks generic.

I want to create an AI-generated image of [describe your basic idea in one sentence — e.g., "a cozy reading nook"]. Help me build a detailed, effective prompt for [Midjourney / DALL-E / Stable Diffusion] by expanding my concept with specific details for each of these categories: (1) Subject — what exactly is in the scene, with specific details, (2) Composition — camera angle, framing, focal point, (3) Lighting — type, direction, quality, color temperature, (4) Style — artistic style, medium, or photographic approach, (5) Mood/Atmosphere — emotional tone and ambiance, (6) Color palette — dominant and accent colors, (7) Technical specifications — resolution, aspect ratio, rendering quality, (8) Style references — artists, photographers, brands, or films with a similar aesthetic. Combine all of these into one cohesive prompt paragraph that I can paste directly into the image generator. Then explain why you chose each element.

💡

Pro tip: Save the expanded prompts as templates. After a few rounds, you will internalize the framework and start writing detailed prompts directly without needing this expansion step.

Lighting and Mood Modifier Generator

32/40

✨ What it does

ChatGPT writes ten lighting descriptions for the same subject, each with a different mood and the exact words to paste. You pick one lighting line and drop it into your next image prompt.

I am generating AI images and I want to master lighting descriptions. For the subject "[describe your subject — e.g., a portrait of a woman, a still life, a cityscape]," generate 10 different versions of a lighting description that each create a completely different mood. For each version, provide: (1) the exact lighting modifier text I should add to my prompt, (2) the mood/emotion it creates, (3) what time of day or setting it implies, and (4) a famous photograph, painting, or film that uses this lighting style as a reference. Range from dramatic and moody to soft and ethereal. Include at least one example each of: natural light, studio light, practical light (from objects in the scene), and mixed lighting.

💡

Pro tip: Lighting is the number one factor that determines whether an AI image looks amateur or professional. Save these lighting modifiers and reuse them across different prompts. The same subject with different lighting is essentially a different image.

Style Transfer Prompt Builder

33/40

✨ What it does

ChatGPT translates the look of a reference artist, film, or era into concrete prompt words for [describe your subject], so an image generator can actually hit that style. You paste the style block in and check that it still looks like your subject.

I want to create an image of [describe your subject] but rendered in the visual style of [reference — e.g., a specific artist, film, video game, design era, or art movement]. Analyze the key visual characteristics of this style and translate them into specific, concrete prompt language that an AI image generator would understand. Break down: (1) color palette — specific colors and relationships, (2) line quality — sharp/soft, thick/thin, visible/invisible, (3) texture and surface — smooth, rough, painted, digital, (4) composition tendencies — symmetric, dynamic, minimal, dense, (5) lighting approach — flat, dramatic, naturalistic, stylized, (6) level of detail — hyperdetailed, simplified, abstracted, and (7) mood and tone — serious, playful, dark, ethereal. Then combine these into a single optimized prompt I can use. Finally, suggest 3 alternative style references that have a similar but distinct aesthetic I might also enjoy.

💡

Pro tip: This technique is especially powerful for styles you can recognize but cannot articulate. You know the look of Wes Anderson films or Moebius comics, but translating that into prompt language is difficult. This prompt does that translation for you.

Negative Prompt and Refinement Assistant

34/40

✨ What it does

ChatGPT reads [paste your original prompt] and the problems you saw, then rewrites the prompt and adds words that push those errors out. You generate again with the new pair and stop when the old flaw is gone.

I generated an AI image using this prompt: "[paste your original prompt]" but the result has these problems: [describe what went wrong — e.g., "the hands look distorted," "the style is too cartoonish," "the background is too busy," "the lighting is flat"]. Help me fix this by: (1) identifying which parts of my prompt likely caused each problem, (2) rewriting the prompt with corrections, (3) suggesting negative prompt terms I should add to explicitly exclude unwanted elements (e.g., "blurry, distorted hands, oversaturated, text, watermark"), (4) recommending specific modifier additions that counteract each issue, and (5) providing the complete revised prompt ready to paste, with your changes highlighted in brackets so I can learn what you modified. If relevant, suggest parameter adjustments (guidance scale, steps, seed) for the specific AI tool I am using.

💡

Pro tip: Always include your failed prompt when asking for help. The AI can diagnose specific issues much better when it can see what you originally wrote. "My image looks bad" is not actionable. "My image looks cartoonish when I wanted photorealism" is.

Batch Prompt Variations Generator

35/40

✨ What it does

ChatGPT rewrites [paste your working prompt] eight times, changing only one big variable each time, such as light or color. You run the set, compare the images, and keep the change that actually moved the look.

Here is my base image prompt: "[paste your working prompt]." I like the results but I want to explore variations. Generate 8 modified versions of this prompt, each changing ONE major variable while keeping everything else the same: (1) change only the lighting to something dramatically different, (2) change only the color palette, (3) change only the camera angle/composition, (4) change only the artistic style/medium, (5) change only the mood/atmosphere, (6) change only the time of day/season, (7) change only the level of detail (more minimal or more complex), (8) change the scale — zoom way in or way out. For each variation, bold the specific words you changed so I can see exactly what was modified. This helps me understand which variables have the biggest impact on the final image.

💡

Pro tip: This is the fastest way to learn prompt engineering. When you change only one variable at a time, you can clearly see its effect. After one round, you will know whether lighting or color palette matters more for your specific subject.

Style Modifiers Library

5 prompts

Camera and Lens Modifier Reference

36/40

✨ What it does

ChatGPT writes camera and lens lines for the same simple subject, from a phone wide shot to a fast portrait lens. You pick one camera line and reuse it on later prompts.

Generate a visual reference guide showing the same subject — [describe a simple subject, e.g., "a ceramic vase with wildflowers on a wooden table"] — as it would appear when photographed with each of these camera/lens combinations: (1) iPhone 16 Pro, wide lens — smartphone aesthetic, deep depth of field, (2) Canon EOS R5 with 85mm f/1.2 — shallow bokeh portrait lens, (3) Hasselblad medium format with 80mm — medium format film look, (4) Fujifilm X100VI — classic rangefinder street photography look, (5) large format 4x5 view camera — tilt-shift, extreme detail. For each, describe specifically how the image changes: depth of field, distortion, color rendering, resolution feel, and overall aesthetic. Then provide the exact prompt modifier text I should append to any prompt to achieve each look.

💡

Pro tip: The camera/lens modifier is one of the most powerful and underused tools in AI image prompting. "Shot on Hasselblad" versus "shot on iPhone" produces dramatically different results even with identical scene descriptions.

Art Medium and Texture Modifiers

37/40

✨ What it does

ChatGPT writes prompt text that renders the same scene as oil, watercolor, and other media, with the words that make each medium read. You pick a medium block and paste it onto your next subject.

Show me how specifying different art media changes the rendering of the same subject. Take this scene: "[describe a simple scene, e.g., a lighthouse on a rocky coast at sunset]" and provide optimized prompt text for rendering it as: (1) oil painting on canvas with visible brushstrokes, (2) watercolor with wet-on-wet bleeding effects, (3) charcoal drawing on textured paper, (4) digital illustration with clean vector lines, (5) woodblock print with limited colors like Japanese ukiyo-e, (6) pencil sketch in a Moleskine notebook, (7) acrylic pour painting with fluid organic patterns, (8) linocut print with bold black and white contrast. For each medium, provide: the exact modifier phrase to add to prompts, what texture and detail characteristics to expect, and one famous artist who works in that medium whose name can be used as an additional style reference.

💡

Pro tip: Combine medium modifiers with artist references for maximum control. "Watercolor in the style of Winslow Homer" produces very different results than "watercolor in the style of Agnes Cecile." The medium sets the technique; the artist reference sets the aesthetic within that technique.

Color Grading and Palette Modifiers

38/40

✨ What it does

ChatGPT writes reusable color-grade lines for a portrait, landscape, or street scene, from warm Portra skin to hard, saturated film looks. You copy one grade into your prompt and generate until the palette matches.

I want to master color control in AI image generation. For a [general scene type — e.g., portrait, landscape, still life, street scene], generate specific prompt modifier text for each of these color approaches: (1) Kodak Portra 400 film emulation — warm skin tones, soft contrast, (2) Fujifilm Velvia — hypersaturated, punchy colors for landscapes, (3) cross-processed film — unexpected color shifts and high contrast, (4) desaturated cinema look — teal and orange color grading, (5) monochromatic — entire image in shades of [single color], (6) complementary color scheme — built on [color] and its complement, (7) analogous palette — [three adjacent colors on the color wheel], (8) high-key pastel — bright, airy, minimal shadows. For each, provide the exact text to add to any prompt and explain when this color approach works best.

💡

Pro tip: Film stock names are the most efficient color modifiers available. "Kodak Portra 400" communicates an entire color science — warm tones, soft contrast, flattering skin rendering — in three words. Learn five film stocks and you have five distinct color palettes at your fingertips.

Composition and Framing Modifiers

39/40

✨ What it does

ChatGPT writes framing lines for the same subject, from rule of thirds to other classic compositions. You pick one framing line, generate, and see where the eye actually lands.

Teach me how to control composition in AI image generation by showing how different framing instructions change the same scene. Subject: "[describe subject — e.g., a woman walking through a field of lavender]." Provide optimized prompt modifier text for each composition technique: (1) rule of thirds — subject at a power point, (2) center-frame symmetry — subject dead center, balanced, (3) leading lines — environmental lines drawing the eye to the subject, (4) frame within a frame — subject seen through a doorway, window, or arch, (5) negative space — subject small in a vast environment, (6) extreme close-up — filling the frame with detail, (7) bird's-eye view — looking directly down, (8) worm's-eye view — looking directly up. For each, provide the prompt text AND explain what emotional effect this composition creates — because composition is not just technical, it communicates feeling.

💡

Pro tip: Composition modifiers have an emotional impact most people overlook. Negative space creates loneliness or freedom. Extreme close-ups create intimacy or claustrophobia. Learning to match composition to emotion is what makes AI images feel intentional rather than random.

Time Period and Era Style Modifiers

40/40

✨ What it does

ChatGPT writes era style lines that drop the same subject into decades from the 1920s onward. You pick one era block, generate, and check that the clothes and design actually belong to that time.

Show me how to transport any subject to different visual eras using AI image prompts. Take this subject: "[describe a neutral subject — e.g., a city street corner, a family portrait, a product]" and provide the exact prompt modifier text to render it authentically in each era: (1) 1920s Art Deco — geometric patterns, gold and black, glamour, (2) 1950s mid-century — atomic age, Googie architecture, pastel optimism, (3) 1970s — warm film grain, earth tones, analog texture, (4) 1980s — neon, grid lines, chrome, synthwave, (5) 1990s — grunge, muted tones, lo-fi, (6) Y2K (2000s) — glossy, metallic, bubbly, tech-optimistic, (7) 2010s minimal — clean, white space, sans-serif, Instagram, (8) retro-futurism — how the past imagined the future. For each era, specify the key visual markers, color palette, typography style, and a cultural reference (film, brand, or designer) that epitomizes the look.

💡

Pro tip: Era modifiers are incredibly useful for marketing and branding. A product shot rendered in "1970s warm film grain" versus "2010s minimal" tells a completely different brand story. Use era modifiers to test how different aesthetic positions feel for your brand before committing to one.

Free tool

Image Prompt Builder

Build a rich copy-paste prompt for Midjourney, DALL-E, or Stable Diffusion.

Try it free →

Frequently Asked Questions

A good AI image prompt has five layers: subject (what is in the image), composition (how it is framed), lighting (what illuminates it), style (what artistic approach defines it), and mood (what emotion it conveys). Most people only specify the subject — "a cat on a rooftop" — and wonder why the results look generic. Adding the other four layers — "a tabby cat sitting on a Parisian rooftop at golden hour, shot from below with a 35mm lens, warm film grain, nostalgic and peaceful" — transforms the output from random to intentional. The most impactful single addition you can make to any prompt is a lighting description. Lighting controls mood more than any other variable.
Each generator has distinct strengths. Midjourney excels at artistic, stylized, and aesthetically beautiful images — it is the best choice for concept art, illustrations, and images where visual beauty matters most. DALL-E (through ChatGPT) is strongest at following complex instructions accurately and handling text in images — use it when prompt adherence and specificity matter more than pure aesthetics. Stable Diffusion offers the most control through its open-source ecosystem of models, LoRAs, and ControlNet — it is ideal for users who want fine-grained technical control and are comfortable with more complex workflows. For beginners, start with DALL-E in ChatGPT for ease of use, then explore Midjourney for higher aesthetic quality.
Distorted anatomy — especially hands — remains a challenge in AI image generation, though models have improved significantly. To minimize issues: (1) use negative prompts explicitly excluding "distorted hands, extra fingers, deformed face, blurry" — this works in Stable Diffusion and Midjourney, (2) avoid poses where hands are the focal point or are performing complex actions, (3) specify a wider framing so hands are smaller in frame and less scrutinized, (4) use inpainting to fix specific areas after generation rather than trying to get everything perfect in one pass, and (5) choose models known for better anatomy — newer model versions consistently handle human anatomy better than older ones. For faces, specifying a real camera lens (like 85mm f/1.4) tends to produce more naturalistic facial rendering.
Commercial usage rights depend on the specific tool and its terms of service, and this area is evolving rapidly. As of 2026: Midjourney allows commercial use for paid subscribers. DALL-E through ChatGPT allows commercial use and gives you full rights to the images you create. Stable Diffusion models vary — most open-source models allow commercial use, but check the specific model license. However, legal and ethical considerations go beyond terms of service. Some jurisdictions are developing regulations around AI-generated content, and copyright questions around AI training data remain unsettled in courts worldwide. For commercial projects, best practices include: clearly disclosing AI-generated content where required, avoiding prompts that specifically mimic a living artist's style, keeping records of your prompts and generation process, and consulting legal counsel for high-stakes commercial applications.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.