From Bad to Great: AI Prompt Before and After Examples

See the Difference: AI Prompt Examples Before and After Optimization

The gap between a mediocre AI-generated image and a stunning one almost always comes down to the prompt — not the model. This guide presents ten real-world AI prompt examples before after transformation, showing exactly how rewording a prompt produces dramatically better results. Each example includes the original weak prompt, the improved version, and a detailed breakdown of why each change matters. Study these transformations, internalize the patterns, and apply them to every prompt you write.

Example 1: Portrait Photography

Before:

“A portrait of a woman with good lighting”

After:

“A medium close-up portrait of a woman in her late 30s with warm brown skin and dark curly hair falling past her shoulders, soft Rembrandt lighting from a large window camera left, gentle closed-lip smile with relaxed eyes, wearing a cream linen blouse, background is a blurred warm-toned interior at f/2.0, shot on 85mm lens, warm 5200K color temperature”

What changed: The “before” prompt contains zero actionable visual information. “Good lighting” means nothing to a model. The “after” prompt specifies the exact framing (medium close-up), subject details (age, skin tone, hair, clothing, expression), lighting pattern (Rembrandt, window source, direction), technical settings (lens, aperture, color temperature), and background treatment. Every element that the model would otherwise decide randomly is now controlled.

Example 2: Product Photography

Before:

“A photo of a perfume bottle”

After:

“A three-quarter angle product photograph of a rectangular glass perfume bottle with amber liquid, catching studio sidelight that creates bright specular highlights along the edges, placed on a sheet of black acrylic creating a mirror reflection below, single eucalyptus branch positioned behind the bottle, deep charcoal background with subtle gradient, shot on 100mm macro lens, f/11 for full sharpness, commercial product photography”

What changed: The before prompt tells the model what to generate but not how. The after prompt specifies the bottle’s physical characteristics (rectangular, glass, amber liquid), the exact lighting effect (studio sidelight, specular highlights on edges), the surface (black acrylic with reflection), styling props, background treatment, and camera specifications. This level of detail produces results that look like professional studio work.

Example 3: Landscape

Before:

“A beautiful mountain landscape”

After:

“A wide-angle landscape of the Dolomites at blue hour, 20 minutes after sunset, jagged limestone peaks silhouetted against a deep violet and indigo sky, a still alpine lake in the foreground reflecting the peaks and sky perfectly, remnants of golden light on the distant ridge, thin layer of ground mist hugging the lake surface, shot on 16mm wide-angle lens, deep depth of field, foreground rocks adding layered depth, long exposure smoothing the water to glass”

What changed: “Beautiful” was replaced with specific visual information — the exact time of day (blue hour, 20 minutes after sunset), named colors (violet, indigo, golden), atmospheric details (ground mist), and compositional elements (reflection, foreground rocks, layered depth). The lens and exposure choices shape the final image feel.

Example 4: Food Photography

Before:

“A photo of pasta”

After:

“An overhead flat-lay photograph of a ceramic bowl of fresh pappardelle pasta with wild mushroom ragù, shaved Parmigiano-Reggiano scattered across the top with fresh thyme sprigs, rustic linen napkin tucked under the bowl on a weathered oak table, soft natural window light from the upper right creating gentle shadows, warm earth tone color palette with pops of green from herbs, food photography styling with scattered breadcrumbs and a vintage brass fork, shot on 50mm lens”

What changed: The subject went from a category (“pasta”) to a specific dish with named ingredients and garnishes. The composition is defined (overhead flat-lay), the surface is described (weathered oak), lighting is specified (window light, upper right), and styling details add authenticity (breadcrumbs, vintage fork, linen napkin). Food photography depends on styling details — the more you specify, the more appetizing the result.

Example 5: Interior Design

Before:

“A modern living room”

After:

“A wide-angle interior photograph of a Scandinavian-minimalist living room, low-profile gray linen sofa against a white plaster wall, a single large abstract painting in muted earth tones above the sofa, pale oak hardwood floors with a natural jute rug, a ceramic vase with dried pampas grass on a minimal wood side table, afternoon sunlight streaming through floor-to-ceiling windows casting long geometric shadows, warm neutral palette of whites, grays, and natural wood tones, interior design magazine photography”

What changed: “Modern” was replaced with a specific design style (Scandinavian-minimalist) that carries clear visual associations. Every furniture piece is described with its material and color. The light source and its effect (geometric shadows from floor-to-ceiling windows) are specified. The color palette is defined in concrete terms. The “interior design magazine photography” style reference tells the model the intended quality level.

Example 6: AI Video — Product Reveal

Before:

“A video of sneakers”

After:

“Camera slowly orbiting 180 degrees around a pair of white leather sneakers on a rotating platform, shoes positioned at eye level against a warm gradient background transitioning from peach to soft coral, dramatic key light from the upper left creating strong shadows that define the shoe’s silhouette, subtle particle dust floating through backlight beams, slow motion rotation revealing stitching details and sole profile, luxury product reveal cinematic style, 4 seconds, 16:9”

What changed: The video prompt adds three motion layers: camera movement (orbiting 180 degrees), subject motion (rotating platform), and environmental motion (floating particles). Duration, aspect ratio, and speed are specified. The lighting is dramatic and directional, supporting the luxury product reveal intent.

Example 7: Fantasy Art

Before:

“A fantasy castle”

After:

“A towering Gothic fantasy castle perched on the edge of a sea cliff, built from dark gray stone with green copper spires, illuminated from within by warm torchlight glowing through narrow windows, a turbulent sea crashing against the cliff base 500 feet below, dramatic storm clouds gathering behind the castle with a single break in the clouds allowing a shaft of golden sunlight to spotlight the highest tower, dark moody palette with the golden light as the focal point, matte painting style with photorealistic detail, epic fantasy book cover composition”

What changed: The generic “fantasy castle” became a specific building with defined materials (dark gray stone, green copper spires), a dramatic setting (sea cliff, 500 feet above turbulent water), atmospheric conditions (storm clouds with a single golden light break), and internal illumination (warm torchlight through windows). The style reference (matte painting, book cover composition) guides the model toward a specific visual quality.

Example 8: Street Photography

Before:

“A city street at night”

After:

“A rain-slicked narrow street in Tokyo’s Shinjuku district at 11pm, neon signs in Japanese characters reflecting as long colored streaks on the wet asphalt, a lone figure with a clear umbrella walking away from camera in the middle distance, steam rising from a ramen shop vent on the left, red and cyan neon dominant colors, shot on 35mm lens at f/1.4 with shallow depth of field, moody cinematic color grading, Blade Runner atmosphere”

What changed: The location became specific (Shinjuku, Tokyo) with cultural details (Japanese neon signs). Weather conditions (rain, wet reflections) add atmosphere. A narrative element (lone figure with umbrella walking away) creates emotional depth. The color palette (red and cyan neon) is defined. A cinematic reference (Blade Runner) anchors the mood. Technical specifications (35mm, f/1.4) shape the visual feel.

Example 9: Minimalist Design

Before:

“A minimalist image”

After:

“A single perfect sphere resting on an infinite white surface, casting a long soft shadow to the right, the sphere has a matte terracotta surface with subtle grain texture, vast negative space surrounding the object occupying only the lower right intersection of the rule of thirds grid, perfectly soft even overhead lighting with no harsh shadows, the composition breathes with 80 percent empty space, fine art minimalist photography”

What changed: “Minimalist” was transformed from a vague aesthetic label into specific compositional instructions — a single object, defined placement (rule of thirds), quantified negative space (80 percent), specified material (matte terracotta with grain), controlled lighting (soft, even, overhead), and shadow behavior (long, soft, rightward). Minimalism in prompts requires maximum specificity about minimal elements.

Example 10: Social Media Content

Before:

“A cool image for Instagram”

After:

“A square 1:1 composition, overhead flat-lay of a marble desk surface with a MacBook keyboard visible at the top edge, a ceramic mug of matcha latte with latte art in the center, scattered succulent leaves and a brass pen, warm natural sidelight from the left creating soft shadows, clean millennial-pink and sage-green color palette, lifestyle content creator aesthetic, bright and airy with high key exposure, Instagram-optimized”

What changed: The aspect ratio is specified (1:1 for Instagram). The composition is defined (overhead flat-lay) with specific objects placed in relation to each other. Colors are named (millennial-pink, sage-green). The lighting is described with direction and quality. The aesthetic is targeted (content creator, bright and airy). Every element serves the platform context.

The Universal Pattern

Across all ten examples, the transformation follows the same pattern:

1. Replace vague adjectives with visual descriptions
2. Add a camera specification — angle, lens, framing
3. Define the lighting — source, direction, quality, color
4. Name the colors — specific color terms, not generic labels
5. Describe textures and materials — what surfaces feel like
6. Set the mood — emotional tone backed by visual details
7. Include composition — where things are in the frame

Use Vidzy’s AI Prompt Generator to apply this pattern automatically, or reference the AI prompt keywords cheat sheet for vocabulary.

FAQ

How much difference does prompt optimization actually make? The difference is transformational, not incremental. A well-optimized prompt produces results that look like they came from a different, more advanced AI model. Most of the quality gap between “average” and “amazing” AI art is in the prompt, not the model. What is the single most impactful change I can make to my prompts? Add lighting direction. Specifying where the light comes from and its quality (soft, hard, warm, cool) has the largest single impact on image quality. “Soft Rembrandt lighting from camera left” alone can transform a flat image into a dimensional photograph. How long should an optimized prompt be? The optimized prompts in this guide average 50 to 80 words. This is the sweet spot — long enough to specify all critical elements, short enough for the model to process coherently. Every word should add unique visual information. Do these before-and-after patterns work for all AI models? Yes. The core principle — replacing vague descriptions with specific visual details — works across DALL-E, Flux, Midjourney, Sora, and Vidzy. The specific keywords may vary slightly, but the transformation pattern is universal. Should I always write long, detailed prompts? Not necessarily. The goal is density, not length. A 30-word prompt packed with specific visual information can outperform a 100-word prompt full of vague quality keywords. Focus on making every word contribute visual meaning.

Transform Your Prompts Today

Every “before” prompt in this guide represents how most people write their first attempt. Every “after” prompt represents a few minutes of thoughtful revision. The skill is learnable, the patterns are repeatable, and the improvement in output quality is immediate. Build optimized prompts with Vidzy’s Prompt Generator, or download Vidzy to generate professional AI prompt examples before after quality directly from your iPhone.