Prompt Library

The 40 Best AI Image Prompts of 2026

40 copy-paste prompts

The standout prompts across every major AI image generator — Midjourney, DALL-E 3, Stable Diffusion, and Sora. Copy, paste, generate. 15 free no-signup, the rest behind a one-tap email.

In short: This page contains 40 copy-paste ready prompts, organized into 4 categories with a description and pro tip for each. The first 15 prompts are free instantly — no signup needed. Hand-curated and tested by the AI Academy team.

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

Best Midjourney Prompts

10 prompts

Golden Hour Cinematic Portrait

1/40

Cinematic portrait of a woman with auburn hair, golden hour backlight, soft warm rim light, shallow depth of field, 85mm f/1.4 lens, Kodak Portra 400 film grain, --ar 4:5 --style raw --v 6

The most reliable portrait formula on Midjourney v6 — backlight + 85mm + Portra 400.

💡

Pro tip: For different moods, only swap the lighting term: "blue hour" for moodier, "overcast diffused" for editorial.

Studio Ghibli Pastoral

2/40

Studio Ghibli style watercolor illustration of a child running through tall grass with a yellow umbrella, golden afternoon sunlight, distant green hills, hand-painted feel, --ar 16:9 --niji 6 --style scenic

Pastoral Ghibli aesthetic — high social-share value, low difficulty to prompt.

💡

Pro tip: `--niji 6 --style scenic` is the magic combo for Ghibli landscapes. Use `--style cute` for Ghibli characters.

Brutalist Architecture B&W

3/40

Imposing brutalist concrete building, harsh midday sunlight casting deep shadows, geometric forms, dramatic low-angle perspective, Helene Binet style architectural photography, black and white, --ar 4:5 --style raw --v 6

High-drama architecture shot — works for portfolio, blog headers, and Instagram architecture accounts.

💡

Pro tip: B&W is the default for brutalism. Color shots tend to feel flat; stick to monochrome.

Minimal Tech Startup Logo

4/40

Minimalist geometric logo design for a tech startup, single bold mark, monochrome, clean sans-serif wordmark below, vector style, --ar 1:1 --no gradient, blur, complexity --v 6

Clean SaaS-style logo — Stripe/Linear aesthetic.

💡

Pro tip: The negative prompt (`--no gradient, blur, complexity`) is non-negotiable. Without it, Midjourney adds noise real logos don't have.

Surreal Floating Islands

5/40

Surreal floating rocky islands suspended over endless cloud sea, warm sunset colors, waterfalls falling from islands into clouds, fantasy concept art, --ar 16:9 --v 6

Fantasy concept art — works for book covers, game backgrounds, desktop wallpapers.

💡

Pro tip: For book-cover style, add "tiny silhouette of a person on island edge" — adds scale and emotional anchor.

Northern Lights Reflection

6/40

Aurora borealis over snowy Norwegian fjord, green and purple lights dancing in sky, perfect reflection in still water below, lone red cabin on shore, long exposure, --ar 16:9 --style raw --v 6

Award-winning landscape composition — aurora + water reflection + foreground anchor.

💡

Pro tip: Aurora photos always need a foreground element (cabin, mountain, person). Pure sky shots feel empty.

Macro Bee on Sunflower

7/40

Extreme macro photograph of a bee on sunflower, water droplets on petals, razor-thin focus plane on the bee's eyes, Canon MP-E 65mm macro lens at 3x magnification, natural sunlight, --ar 3:2 --style raw --v 6

Nature-magazine-quality macro — strong for nature blogs and educational content.

💡

Pro tip: The "MP-E 65mm at 3x" reference is what triggers macro-specific lens behavior in Midjourney. Memorize it.

Cyberpunk Anime Character

8/40

Cyberpunk anime character portrait, female warrior with neon pink hair, glowing cybernetic eye implant, leather jacket with chrome details, neon-lit Tokyo alley background, Akira-inspired, --ar 2:3 --niji 6 --style expressive

High-energy cyberpunk character — popular for game/manga moodboards.

💡

Pro tip: Add `--stylize 750` for more dramatic interpretation, or `--stylize 250` for closer adherence to your descriptors.

Cozy Reading Nook

9/40

Cozy reading nook interior, built-in bookshelves, large window seat with linen cushions, warm Edison bulb pendant light, autumn afternoon light, hardwood floors, Architectural Digest style, --ar 4:5 --style raw --v 6

Pinterest-viral interior shot — combines all four virality triggers (cozy + warm light + wood + accent objects).

💡

Pro tip: Always include warm light + natural wood + 1-2 accent objects (coffee cup, open book). That's the formula.

Tokyo Street at Night

10/40

Street photography of rainy Tokyo Shinjuku at night, neon signs reflecting in wet pavement, lone figure with umbrella in foreground, Fujifilm X100V 23mm f/2, Blade Runner 2049 color grade, --ar 16:9 --style raw --v 6

Cinematic urban scene with the wet-neon aesthetic loved by photo communities.

💡

Pro tip: The "Blade Runner 2049 color grade" reference is potent. Other strong references: "Wes Anderson palette", "Roger Deakins cinematography".

Prompts get you started. Tutorials level you up.

A growing library of 300+ hands-on AI tutorials. New tutorials added every week.

Start 7-Day Free Trial

Best DALL-E 3 Prompts

10 prompts

Whimsical Children's Book Illustration

11/40

A whimsical children's book illustration of a small white rabbit reading a giant book in a cozy mushroom library, soft pastel colors, watercolor texture, warm afternoon light through circular windows, in the style of Beatrix Potter and Mary Blair.

DALL-E 3 excels at literary illustration. The "in the style of [named illustrators]" pattern is its strongest control.

💡

Pro tip: DALL-E 3 reads natural language. Write full sentences, not Midjourney-style comma lists. It rewards descriptive prose.

Realistic Product Photography

12/40

A photorealistic product photograph of a matte black ceramic coffee mug on a white seamless backdrop, single softbox light from upper-left creating soft shadow on the right, subtle reflection on the surface below the mug, ultra-clean e-commerce photography style.

DALL-E 3 produces excellent product shots. The trick is being explicit about lighting direction and surface treatment.

💡

Pro tip: For luxury product shots, swap "white seamless backdrop" to "polished marble surface with directional spotlight".

Vintage Travel Poster

13/40

A vintage 1920s travel poster advertising Paris, featuring an Art Deco illustration of the Eiffel Tower at sunset with bold geometric clouds, limited muted color palette of dusty rose and gold, decorative typography reading "VISIT PARIS" at the top.

DALL-E 3 nails vintage poster style — Art Deco, mid-century travel, propaganda posters.

💡

Pro tip: For other eras, change the decade: "1950s mid-century travel poster" feels totally different from "1920s Art Deco".

Educational Diagram

14/40

A clear educational diagram of the water cycle drawn in a friendly, hand-illustrated style, showing evaporation from ocean, cloud formation, rainfall over mountains, and rivers flowing back to ocean. Labels in clean handwritten font. Pastel blue and green color scheme. Suitable for a 5th-grade science textbook.

DALL-E 3 is uniquely good at diagrams with readable labels — Midjourney struggles with this.

💡

Pro tip: For diagrams, always specify "labels in [font style]" and a target audience ("for 5th-grade textbook"). DALL-E renders labels much better than Midjourney.

Realistic Food Photography

15/40

An overhead photograph of fresh handmade pasta with truffle shavings on a rustic white ceramic plate, photographed in natural soft window light from the left, sitting on a weathered wooden table, with a single sprig of parsley and a small dish of olive oil. Restaurant menu photography style.

DALL-E 3 nails food photography when you specify camera angle, lighting direction, and surface explicitly.

💡

Pro tip: Overhead works for plated dishes; "45-degree angle" works for drinks and layered desserts.

Brand Mascot Character

16/40

A friendly cartoon mascot character for a children's educational app: a smiling cartoon owl wearing round glasses and holding a small book. Bright primary colors, thick black outlines, simple geometric shapes, sticker-style flat illustration. Single character, white background, designed for use as an app icon.

DALL-E 3 produces clean mascot characters with consistent style — strong for brand identity work.

💡

Pro tip: Specify "single character, white background" to get usable assets vs. illustrated scenes.

Infographic Hero Image

17/40

A modern flat-illustration hero image for a productivity blog article titled "10 Ways to Beat Procrastination". A person at a clean desk with a laptop, surrounded by floating icons of clocks, checkmarks, and lightbulbs. Vibrant orange and teal color scheme, isometric perspective, cheerful but professional tone.

Editorial illustration style — Notion / Linear / Stripe blog post header energy.

💡

Pro tip: For other moods: "muted earth tones + flat illustration" reads thoughtful; "bright primary + isometric" reads energetic.

Holiday Greeting Card

18/40

A warm winter holiday greeting card illustration: a cozy red brick cottage covered in snow with warm yellow light glowing from windows, smoke rising from chimney, decorated pine wreath on door, soft snowfall, in the style of Norman Rockwell. The words "Happy Holidays" in elegant gold script across the top.

Holiday card style — works for Christmas, Diwali, Hanukkah scenes.

💡

Pro tip: DALL-E 3 handles short text overlays better than Midjourney. Keep text under 4 words for best results.

Logo Concept with Type

19/40

A modern logo concept for a coffee subscription brand called "Morning Ritual". A minimal line-drawing of a steaming coffee cup with a sun rising behind it. Brand name "MORNING RITUAL" set in clean modern sans-serif beneath the icon. Color palette: warm cream and deep espresso brown.

DALL-E 3 can produce passable logos with readable text — Midjourney almost never can.

💡

Pro tip: Always specify the brand name in quotes and the font style. DALL-E renders short brand names accurately ~70% of the time.

Stylized Map Illustration

20/40

A whimsical hand-drawn illustrated map of a fantasy island in the style of a Hobbit book endpaper: rolling hills, a forest in the north, a tiny village with cottages, a winding river, and a small castle. Labeled in vintage script. Sepia tones on aged parchment. Decorative compass rose in the corner.

Stylized maps for fantasy projects, RPGs, book endpapers, or quirky branding.

💡

Pro tip: "Aged parchment" + "vintage script labels" is the unlock combo. Try "watercolor on cotton paper" for a softer feel.

Best Stable Diffusion Prompts

10 prompts

Photorealistic Portrait — SDXL

21/40

RAW photo, portrait of a young woman, freckles, natural makeup, soft window light from left, 85mm lens, f/1.4, shallow depth of field, kodak portra 400, film grain, photographed by Annie Leibovitz, masterpiece, ultra-detailed, 8k. Negative prompt: cartoon, 3d render, plastic skin, oversaturated, blurry, low quality

SDXL portrait formula — note the explicit negative prompt, which Stable Diffusion needs more than Midjourney does.

💡

Pro tip: Stable Diffusion REQUIRES negative prompts to avoid common failure modes (plastic skin, oversaturation). Always include the negative.

Anime Character — SDXL Anime Model

22/40

1girl, long flowing silver hair, blue eyes, school uniform, cherry blossom petals falling, soft pastel color palette, anime style, masterpiece, best quality, highly detailed, by makoto shinkai. Negative prompt: lowres, bad anatomy, bad hands, missing fingers, blurry, watermark

Anime character on an SDXL anime model (NAI, AnythingV5, etc.) — uses tag-style prompting.

💡

Pro tip: SD anime models use tag-style prompts (comma-separated keywords), not natural language. Order matters — leading tags carry more weight.

Concept Art — Fantasy Environment

23/40

Epic fantasy concept art of an ancient elven city in the mountains, glowing crystal towers, waterfalls, dragons flying in distance, golden hour, cinematic composition, by Ralph McQuarrie and Greg Rutkowski, ArtStation trending, highly detailed, 8k. Negative prompt: cartoon, low quality, blurry, watermark, signature

Game/film concept art style — Greg Rutkowski reference is one of SD's strongest style anchors.

💡

Pro tip: Stacking multiple named artists ("by X and Y and Z") blends styles. Up to 3 names works; more gets noisy.

Product Render — Industrial Design

24/40

Professional product render of a futuristic ergonomic computer mouse, matte titanium finish, soft studio lighting, white seamless backdrop, octane render, blender, hyper-detailed, industrial design portfolio. Negative prompt: hand, sketch, drawing, painting, blurry, distorted

Industrial-design product render — Behance / Dribbble aesthetic.

💡

Pro tip: "Octane render" + "blender" trigger the CGI-quality look. For softer plastic feel, swap to "Keyshot render".

Architectural Visualization

25/40

Architectural visualization of a modern lakeside house, large glass facade, warm interior lighting at dusk, surrounded by pine trees, lake reflection in foreground, vray render, ultra-detailed, by Iwan Baan. Negative prompt: painting, sketch, cartoon, lowres

Architectural visualization (archviz) — used by architects for client presentations.

💡

Pro tip: "Vray render" is the industry term — it produces more believable architecture than "octane render" for buildings.

Album Cover Art

26/40

Atmospheric album cover art for an indie folk band, lone figure walking through misty pine forest at dawn, muted earth tones, painterly texture, by Andrew Wyeth and Caspar David Friedrich, melancholy mood, square format. Negative prompt: text, watermark, signature, cartoon

Album-cover aesthetic — Wyeth + Friedrich combo unlocks the painterly-melancholy mood that's popular in indie folk.

💡

Pro tip: Specify "square format" for album covers. For vinyl LP feel, add "vintage texture, slight film grain, faded colors".

Vintage Movie Poster

27/40

Vintage 1970s science fiction movie poster, lone astronaut in front of giant alien planet, painted illustration by Drew Struzan, retro typography, faded color palette, vertical poster format. Negative prompt: photograph, modern, digital, cartoon

Retro movie poster style — Star Wars / Indiana Jones / Goonies poster aesthetic via Drew Struzan reference.

💡

Pro tip: Naming "Drew Struzan" specifically triggers the 70s/80s movie poster look. For newer styles, name modern poster artists.

Watercolor Botanical

28/40

Delicate watercolor botanical illustration of monstera deliciosa plant, soft greens, white background, hand-painted feel, by Marianne North, scientific illustration style. Negative prompt: photograph, 3d, digital, cartoon, harsh lines

Botanical watercolor — wedding invitations, journal pages, gallery prints.

💡

Pro tip: "By Marianne North" anchors scientific botanical illustration. For modern style, try "by Katie Brown" or "modern minimalist watercolor".

Pixel Art Game Asset

29/40

16-bit pixel art sprite of a medieval knight character, sword and shield, idle pose, 32x32 pixel resolution upscaled, retro RPG style, in the style of Chrono Trigger and Stardew Valley. Negative prompt: 3d, photorealistic, blurry, anti-aliased

Pixel art game assets — usable for indie game development or retro aesthetic projects.

💡

Pro tip: "Chrono Trigger and Stardew Valley" combo is the gold standard for SNES-era pixel art. Specify pixel resolution explicitly.

Cyberpunk Cityscape

30/40

Cyberpunk megacity at night, towering skyscrapers with massive neon signs in Japanese and English, flying cars between buildings, rain-soaked streets reflecting neon, dense atmosphere, by Syd Mead and Moebius. Negative prompt: daytime, clear sky, empty, low quality

Iconic cyberpunk cityscape — Blade Runner / Ghost in the Shell aesthetic via Syd Mead.

💡

Pro tip: "Syd Mead" is the original cyberpunk concept artist (Blade Runner, Tron). Combining with "Moebius" adds the European comic style.

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

Best Sora Prompts (Video)

10 prompts

Slow-Motion Wave Crash

31/40

A massive ocean wave crashing against jagged cliffs in slow motion, shot at 240 frames per second, golden hour backlight catching sea spray in mid-air, ultra-wide cinematic shot, IMAX-quality detail, dramatic and powerful atmosphere. Duration: 8 seconds.

Showcase slow-motion clip — works as standalone art piece or transition.

💡

Pro tip: Sora handles slow motion well when you specify the framerate. "240fps" or "high-speed cinematography" both work.

Cinematic City Walk

32/40

Steadicam tracking shot following behind a woman in a red coat walking through a rainy Paris evening, neon shop reflections in wet cobblestones, soft background bokeh, shot on Arri Alexa, color graded like the film Amelie. Duration: 10 seconds.

Cinematic tracking shot — feels like a scene from a feature film.

💡

Pro tip: "Arri Alexa" + "color graded like [film]" is a strong cinematography combo. Substitute different films for different moods.

Aerial Drone Landscape

33/40

Slow aerial drone shot rising up from a misty pine forest at dawn, revealing mountain peaks above the fog, soft golden sunrise light, smooth gimbal movement, ultra-cinematic, Planet Earth documentary style. Duration: 10 seconds.

Establishing aerial shot — strong for travel content, film openers, or atmospheric video.

💡

Pro tip: "Planet Earth documentary style" anchors the smooth, deliberate camera movement and natural color grading.

Product Reveal — Macro

34/40

Smooth macro dolly shot circling around a luxury wristwatch on a black velvet surface, soft moving spotlight revealing different details, dust motes visible in light, premium product photography aesthetic. Duration: 8 seconds.

Luxury product reveal — usable for watch, jewelry, or premium tech product launches.

💡

Pro tip: Specify "circling dolly shot" vs "static" — Sora needs explicit camera movement direction.

Time-Lapse — City Sunset

35/40

Time-lapse of city skyline transitioning from sunset to night, sun setting on left side of frame, building lights gradually turning on, clouds moving across sky, traffic light trails on streets below, smooth motion. Duration: 10 seconds.

Time-lapse cityscape — works as B-roll or social-media establishing shot.

💡

Pro tip: Specify what should change over time ("lights turning on", "clouds moving"). Vague time-lapses come out as static images.

Anime-Style Action Scene

36/40

Anime-style action scene, young swordsman with flowing red cape leaping toward camera from rocky cliff at sunset, dynamic motion blur, dramatic anime lighting, in the style of Studio Trigger. Duration: 6 seconds.

Anime action sequence — usable for game trailers, animated content, or moodboards.

💡

Pro tip: "Studio Trigger" anchors high-energy anime. For more contemplative anime motion, use "Makoto Shinkai" or "Studio Ghibli".

Cozy Cafe Interior — Ambient

37/40

Slow cinematic pan across a cozy cafe interior in autumn, steam rising from coffee cups, soft jazz mood, warm pendant lighting, rain visible through window, single barista working in background, lo-fi aesthetic. Duration: 12 seconds.

Ambient/lo-fi atmosphere clip — great for cafe content, study videos, or social-media B-roll.

💡

Pro tip: "Lo-fi aesthetic" is shorthand for warm tones + slow motion + cozy details. Pair with specific actions (barista working) to ground the scene.

Fashion Editorial Motion

38/40

Slow-motion fashion editorial clip, model walking toward camera through a brutalist concrete corridor wearing a flowing white dress, harsh top lighting, dramatic shadows, monochrome color palette, in the style of a luxury brand campaign. Duration: 8 seconds.

Fashion campaign motion — luxury brand aesthetic with brutalist setting.

💡

Pro tip: "Brutalist concrete corridor" + "monochrome" is the current luxury fashion video formula (Bottega Veneta, The Row, etc.).

Wildlife Documentary Shot

39/40

Slow tracking shot of an Arctic fox walking through fresh snow at sunrise, soft pink and orange dawn light, breath visible in cold air, snow crystals catching light, Planet Earth-quality cinematography, telephoto lens compression. Duration: 10 seconds.

Wildlife documentary — usable for nature content or as standalone art.

💡

Pro tip: "Telephoto lens compression" gives the flattened, intimate wildlife look. Without it, the perspective feels too wide.

Surreal Levitation

40/40

A woman in flowing white dress gently levitating above a misty lake at dawn, hair floating softly upward, dress billowing in slow motion, ethereal lighting, dreamlike atmosphere, in the style of Tim Walker photography brought to motion. Duration: 8 seconds.

Surreal art piece — works as gallery video, music video B-roll, or shareable Instagram content.

💡

Pro tip: Surreal motion clips need ONE impossible element + photoreal everything else. Too much surrealism reads as glitchy.

Frequently Asked Questions

Different generators excel at different things. **Midjourney**: best for artistic, painterly, cinematic images and consistent style. **DALL-E 3**: best for natural-language prompts, accurate text in images, diagrams, and short copy. **Stable Diffusion**: best when you need control via custom models, ControlNet, or local generation. **Sora**: the current leader for short AI video clips. Most pros use a combination — Midjourney for hero art, DALL-E for assets with text, SD for custom workflows.
Yes — every prompt on this page is free to copy and paste. The first 15 are visible immediately with no signup. Drop your email to unlock the rest plus a 10-page guide on the prompt-writing framework.
Yes, but expect different results. Midjourney prompts use comma-separated descriptors and explicit parameters (`--ar`, `--v 6`); DALL-E reads natural-language sentences. A Midjourney prompt pasted into DALL-E often produces something decent but generic; same in reverse. For best results, restructure the prompt to match the model's preferred format.
Three usual reasons: (1) AI image models are stochastic — same prompt produces different images every run. Generate 4+ variations. (2) Your account settings affect every prompt. Check `/settings` in Midjourney, model selection in SD/DALL-E. (3) Small wording matters more than you'd think. "Soft warm light" vs "soft golden light" produces visibly different images. Use exact wording from prompts that work, then iterate.
Depends entirely on the generator and your subscription tier. Midjourney: commercial use requires Pro or higher; Basic plan is limited. DALL-E 3 (via ChatGPT Plus): you own the rights to images you generate but OpenAI retains some usage rights. Stable Diffusion: open license, fully commercial. Sora: commercial-use rights vary by tier. Always check the current terms — they change.
Three habits: (1) Include lighting + camera/lens descriptors — even fake ones work because the model treats them as style hints. (2) Reference named artists, photographers, or films — "Roger Deakins cinematography", "Tim Walker style", "Wes Anderson palette". These compress a whole aesthetic into one phrase. (3) Iterate on ONE element at a time. Change only the lighting in 4 variations, then only the lens, then only the color palette. This is how you build intuition for what each variable controls.

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.

Start 7-Day Free TrialMaster AI image generation at AI Academy

7-day free trial. Cancel anytime.

Browse All Prompt Collections