How to Write AI Image Prompts: The 6-Part Formula (with 25+ Examples)

Why your AI images look generic (and how to fix it)

Two people open the same AI image generator. One types a cool dragon and gets something forgettable. The other writes a detailed brief and gets a jaw-dropping, cinematic result. The difference isn't the tool or luck — it's the prompt.

A great AI image prompt works like a creative brief for a photographer or illustrator. The more clearly you describe what you want, the closer the output lands. This guide breaks that brief into a repeatable 6-part formula, with 25+ copy-paste examples you can adapt. Every example is built to work in Picturagen's generator, where these six elements map directly to on-screen controls.

Quick answer: A strong AI image prompt names six things — subject, style, lighting, composition, mood/color, and format (aspect ratio + quality). Describe each one specifically and you'll rarely get a generic result.

The 6-part AI image prompt formula

Strong 2026 prompts share the same anatomy. Fill each slot and you'll rarely get a generic result:

1. Subject — who or what is in the image (be specific).
2. Style — the visual language (photorealistic, anime, oil painting, 3D…).
3. Lighting — how it's lit (golden hour, cinematic, neon, soft studio…).
4. Composition — framing and camera (close-up portrait, wide shot, top-down…).
5. Mood / color — the feeling and palette (moody teal-and-orange, pastel, high-contrast…).
6. Format / technical — aspect ratio and quality (9:16, 4:5, highly detailed, 8K…).

The formula in one line

[Subject] + [Style] + [Lighting] + [Composition] + [Mood/Color] + [Format]

Weak prompt: a woman in a city

Strong prompt: a confident woman in a tailored charcoal coat walking through a rainy Tokyo street at night, photorealistic, neon signage reflected in wet pavement, cinematic medium shot from slightly below, moody teal-and-magenta palette, highly detailed, 4:5 portrait

Same subject — completely different result.

Slot-by-slot, with examples

1. Subject: be specific

Vague subjects produce vague images. Add age, clothing, action, and defining details.

• Weak: a knight
• Strong: a battle-worn female knight in dented silver armor, kneeling, gripping a cracked longsword

2. Style: pick a visual language

Style is the biggest lever on how the image feels. Picturagen's generator includes ready styles you can name directly in the prompt:

Realistic · Digital Art · Anime / Manga · Cartoon / Comic · Cyberpunk · Fantasy · Oil Painting · 3D Rendered · Pixel Art · Watercolor / Ink · Minimal / Flat

Example: a fox in a forest, watercolor and ink style, loose brushwork

Not sure which style to pick? Our 50 AI art style ideas has a copy-paste example for every one.

3. Lighting: set the mood with light

Lighting separates amateur from professional-looking results.

• golden hour, warm rim light — flattering, cinematic
• soft studio softbox lighting — clean, professional portraits
• neon glow, high contrast — cyberpunk energy
• overcast diffused light — even, editorial

Example: portrait of an elderly fisherman, photorealistic, golden hour rim light

4. Composition: direct the camera

Tell the model where the "camera" is.

• extreme close-up, medium shot, full-body wide shot
• top-down flat lay, low-angle hero shot, over-the-shoulder
• rule of thirds, centered symmetrical composition

Example: a steaming bowl of ramen, top-down flat lay, centered composition

5. Mood & color: choose a palette

• moody teal-and-orange, soft pastel palette, monochrome black and white
• warm autumn tones, high-key bright and airy, dark and dramatic

Example: abstract geometric shapes, minimal flat style, soft pastel palette

6. Format & technical: match where it'll live

Pick the aspect ratio for the destination (Picturagen offers square, 4:5 portrait, landscape, widescreen, and 9:16 story):

• 1:1 square — profile pictures, feed posts
• 4:5 portrait — Instagram feed
• 9:16 — Reels, Stories, TikTok
• 16:9 widescreen — banners, thumbnails

Add quality boosters like highly detailed, 8K, sharp focus where appropriate. For the exact ratio and pixel size to use per platform, see our guide to AI image sizes for social media.

Negative prompts: tell the AI what to avoid

Negative prompts remove common artifacts and are especially powerful for realism:

• For photorealism: blurry, low quality, watermark, text, deformed, extra limbs, bad anatomy
• For portraits: plastic skin, airbrushed, oversaturation, unnatural lighting
• For landscapes: floating objects, blurry edges, overexposed highlights

Model-specific tips

Different engines like different prompt "dialects":

• DALL·E 3 rewards natural, descriptive sentences and handles text-in-image well.
• Midjourney rewards concise, comma-separated descriptors plus parameters.
• Stable Diffusion / Flux reward weighted keywords and strong negative prompts.

The good news: the 6-part formula works across all of them. Start descriptive, then adjust one variable at a time.

25+ copy-paste prompt examples

Portraits & people

1. professional headshot of a woman, photorealistic, soft studio lighting, medium shot, neutral charcoal background, 4:5
2. elderly craftsman in his workshop, photorealistic, warm window light, environmental portrait, muted earth tones
3. cyberpunk street samurai, neon rim light, low-angle hero shot, teal-and-magenta, highly detailed
4. anime schoolgirl under cherry blossoms, soft pastel palette, medium shot, gentle spring light
5. fantasy elf ranger, oil painting style, golden hour, full-body, forest backdrop

Landscapes & environments

1. misty mountain lake at dawn, photorealistic, soft diffused light, wide shot, cool blue palette, 16:9
2. neon-soaked Tokyo alley at night, cinematic, reflections on wet pavement, moody, 9:16
3. rolling Tuscan hills, oil painting, golden hour, warm autumn tones
4. alien desert with twin suns, digital art, dramatic backlight, epic wide shot
5. cozy cabin interior in winter, 3D rendered, warm firelight, soft focus

Products & mockups

1. matte black wireless headphones, product shot, soft studio softbox, centered, minimal flat background, 1:1
2. glass perfume bottle, photorealistic, dramatic side light, high contrast, dark background
3. stack of pancakes with syrup, top-down flat lay, bright natural light, warm tones
4. minimalist app UI mockup, flat design, soft shadows, clean, 16:9
5. artisan coffee cup on wood table, photorealistic, morning window light, shallow depth of field

Creatures & characters

1. majestic dragon over a stormy sea, digital art, dramatic lightning, epic wide shot, cool palette
2. cute robot companion, 3D rendered, soft studio light, centered, pastel colors
3. mythical phoenix rising, oil painting, fiery golden light, dynamic composition
4. pixel-art knight sprite, retro 8-bit style, flat lighting, centered
5. watercolor koi fish, loose ink brushwork, soft palette, minimal background

Abstract & design

1. flowing liquid gold ribbons, 3D rendered, studio light, black background, luxurious mood
2. geometric bauhaus poster, minimal flat, bold primary colors, centered, 4:5
3. dreamy nebula in deep space, digital art, vivid purples and blues, wide
4. paper-cut layered mountains, minimal flat, pastel gradient, 16:9
5. macro shot of a dew-covered spiderweb, photorealistic, soft morning light

Mix and match the six slots to generate dozens more — swap the style or lighting on any line above for an instant variation.

Put the formula to work

The fastest way to internalize this is to generate. Try one prompt, then change a single slot — swap oil painting for cyberpunk, or golden hour for neon glow — and watch how the image transforms. For inspiration, browse real prompts and results in the Picturagen public gallery.

Sources

[1] SurePrompts — AI Image Prompting: The Complete 2026 Guide. [2] SurePrompts — How to Write AI Image Prompts: The 6-Part Formula (2026). [4] BluminAI — AI Image Prompt Guide (2026). [7] Industry consensus on multi-slot prompt anatomy (2026).