1. 3D Collectible Figurine

The viral classic. Try this remix idea:

a 1/7 scale collectible figurine of the person in this photo, displayed on a desk, acrylic base, box packaging visible in the background, studio lighting

2. Retro 90s Bollywood / Hollywood Portrait

Try this remix idea:

Transform this photo into a 1990s film poster portrait — grainy texture, dramatic studio backdrop, vintage color grading, retro typography

3. Anime Character

Try this remix idea:

Redraw the person in this photo as an anime character, cel-shaded, expressive eyes, detailed hair, keeping the same pose and outfit

4. Renaissance Oil Painting

Try this remix idea:

A classical Renaissance oil painting of the person in this photo, chiaroscuro lighting, ornate gold frame, museum quality

5. Cinematic Movie Poster

Try this remix idea:

Turn this photo into an epic action movie poster — dramatic rim lighting, lens flare, title typography, moody teal-and-orange grade

6. Pixar-Style 3D Character

Try this remix idea:

Recreate the person as a Pixar-style 3D animated character, soft global illumination, oversized expressive eyes, friendly proportions

7. Cyberpunk Neon Portrait

Try this remix idea:

The person in this photo as a cyberpunk character — neon rim lighting, holographic HUD elements, rain-slicked city bokeh background

8. LEGO / Toy Brick Minifigure

Try this remix idea:

Convert the person into a toy brick minifigure inside its retail box, product photography, bright studio lighting

9. Polaroid With a Celebrity-Style Caption

Try this remix idea:

A vintage instant photo of the person, flash photography look, white frame, handwritten caption at the bottom, slightly faded colors

10. Fantasy RPG Character Card

Try this remix idea:

The person as a fantasy RPG character — ornate armor, glowing runes, painted card-game art style, stat-card border framing

How to Run These Remixes

Multimodal tools like Gemini and ChatGPT accept an uploaded photo plus a text prompt directly. For the best results:

• Use a well-lit, front-facing photo with a clean background.
• Keep the identity instruction explicit: “keep the same face and features.”
• Change one style element at a time if the result drifts.

The AI Prompt Book Android app makes this workflow one tap: browse the remix category, pick a style you like from the preview images, hit Remix with Photo, and choose a photo from your gallery. The app pairs your photo with the tested prompt so you skip the trial-and-error.

New to prompting? Start with our beginner’s guide to writing AI image prompts.

The Anatomy of a Great Prompt

Every strong image prompt answers four questions, in roughly this order:

1. Subject — Who or what is in the image? Be specific: “a weathered lighthouse keeper in his 60s” beats “an old man.”
2. Setting & Action — Where are they, and what’s happening? “standing on a cliff edge at dawn, holding a brass lantern.”
3. Style & Medium — How should it look? “oil painting,” “35mm film photograph,” “Studio Ghibli style,” “3D render.”
4. Technical modifiers — Lighting, composition, quality: “golden hour, shallow depth of field, highly detailed.”

Put together:

A weathered lighthouse keeper in his 60s standing on a cliff edge at dawn, holding a brass lantern, 35mm film photograph, golden hour lighting, shallow depth of field, highly detailed

Style Keywords Do the Heavy Lifting

The single fastest way to improve your results is a strong style vocabulary. A few reliable families:

• Photography: “portrait photography,” “macro shot,” “drone view,” “long exposure”
• Art movements: “impressionist,” “art deco,” “ukiyo-e,” “vaporwave”
• Render engines: “Unreal Engine 5,” “octane render,” “isometric 3D”
• Artists & studios: “in the style of Studio Ghibli,” “Pixar style character”

Don’t stack ten styles in one prompt — pick one dominant style and one or two supporting modifiers.

Negative Prompts: Say What You Don’t Want

Stable Diffusion and several other tools support negative prompts — a second field listing things to avoid: blurry, extra fingers, watermark, low quality, deformed hands. In Midjourney, use the --no parameter: --no text, watermark.

Common Beginner Mistakes

• Too vague. “Beautiful landscape” gives the model nothing to work with.
• Too long. Past ~60 words, most models start ignoring details. Prioritize.
• Contradictions. “Minimalist detailed baroque” confuses the model. Pick a lane.
• Ignoring aspect ratio. A phone wallpaper needs --ar 9:16; a banner needs --ar 16:9.

The Shortcut: Start From Tested Prompts

The fastest way to learn prompt structure is to study prompts that already work. That’s exactly what the AI Prompt Book app is for — hundreds of tested prompts organized by style, each with a preview image. Copy one, run it, then tweak one element at a time (subject, lighting, style) and watch how the output changes. You’ll internalize the structure in a week of casual experimenting.

Once you’re comfortable, dive into advanced techniques like cinematic lighting control.

1. The “Volumetric” Secret

Most users stop at “soft lighting.” To create true environmental depth, use specific physics-based tokens like “volumetric lighting” or “Tyndall effect.” This instructs the model to simulate light passing through atmospheric particles (dust, fog, or steam), creating those iconic God rays that give a scene instant scale and drama.

Try this prompt: /imagine prompt: A cyberpunk monk in a rain-slicked alley, volumetric neon lighting, Tyndall effect, cinematic chiaroscuro, 8k resolution, photorealistic --v 6.0

2. Color Temperature Blending

Cinematic visuals rely heavily on contrast between warm and cool tones to direct the viewer’s focus. Instead of listing generic colors, try the “Golden Hour vs. Blue Hour” technique: explicitly define two competing color temperatures as distinct light sources. This creates a dynamic tension that mimics real-world cinematic setups.

3. The Material Reflection Hack

Light is only as good as the surface it hits. Instructing the AI on specific textures like “anodized aluminum,” “wet obsidian,” or “translucent polycarbonate” forces the renderer to calculate complex ray-tracing interactions, giving your lighting a tactile, high-tech finish.

Premium workflow: /imagine prompt: Close up of a futuristic mechanical heart, bioluminescent core, reflections on wet obsidian, high-key studio lighting, sharp focus --ar 16:9 --style raw

4. Name the Light Source, Not the Effect

“Dramatic lighting” is vague. “A single sodium-vapor street lamp overhead” is a physical instruction the model can render consistently. Practical sources — candles, neon signage, monitor glow, moonlight through blinds — anchor your scene in believable physics.

5. Chiaroscuro for Instant Mood

Borrow from classical painting: the keyword “chiaroscuro” pushes extreme light-dark contrast that sculpts faces and objects dramatically. Pair it with “low-key lighting” for noir moods, or “Rembrandt lighting” for portrait work.

By mastering these lighting techniques, you move away from generic generations toward intentional, art-directed masterpieces. The prompt is the brush, but the lighting is the paint — experiment with blending these hacks in your next session.