137 lines
6.4 KiB
Markdown
137 lines
6.4 KiB
Markdown
# Gemini Image Generation Prompt Guide
|
|
|
|
Official prompting strategies from Google's Gemini documentation.
|
|
|
|
## Core Principle
|
|
|
|
**"Describe the scene, don't just list keywords."**
|
|
|
|
The model excels at processing narrative descriptions rather than disconnected terms. Comprehensive prompts are significantly more successful than keyword lists.
|
|
|
|
## Seven Effective Prompt Strategies
|
|
|
|
### 1. Photorealistic Scenes
|
|
|
|
Use photography terminology to guide realistic outputs. Reference camera angles, lens specifications, lighting conditions, and fine details.
|
|
|
|
**Key elements**: Shot type, subject, action/expression, environment, lighting description, mood, camera/lens details, textures, aspect ratio.
|
|
|
|
**Example approach**: Specify an "85mm portrait lens" for soft bokeh, "golden hour light" for warmth, and describe specific textures like "sun-etched wrinkles."
|
|
|
|
### 2. Stylized Illustrations & Stickers
|
|
|
|
Be explicit about artistic style and request specific visual treatments like transparent backgrounds or cel-shading.
|
|
|
|
**Key elements**: Style category, subject characteristics, line style, shading approach, color palette, background specifications.
|
|
|
|
**Example**: "kawaii-style sticker" with "bold, clean outlines" and "simple cel-shading."
|
|
|
|
### 3. Accurate Text in Images
|
|
|
|
Gemini excels at rendering readable text. Clearly specify the text content, font style (descriptively), and design aesthetic.
|
|
|
|
**Key elements**: Image type, brand/concept, exact text, font style description, design style, color scheme.
|
|
|
|
### 4. Product Mockups & Commercial Photography
|
|
|
|
Create professional product shots for ecommerce using studio lighting terminology.
|
|
|
|
**Key elements**: Studio-lit photography setup, surface/background, three-point lighting setup, camera angle, sharp focus areas, realism level, aspect ratio.
|
|
|
|
### 5. Minimalist & Negative Space Design
|
|
|
|
Perfect for backgrounds where text will overlay. Position a single subject with vast empty space.
|
|
|
|
**Key elements**: Singular subject, placement position, background color, negative space amount, lighting direction, aspect ratio.
|
|
|
|
### 6. Sequential Art (Comic/Storyboard)
|
|
|
|
Build on character consistency and scene description for visual storytelling.
|
|
|
|
**Key elements**: Panel count, artistic style, character placement, scene type, specific visual qualities.
|
|
|
|
### 7. Grounding with Google Search
|
|
|
|
Generate images based on current information using real-time data for news, weather, and time-sensitive topics.
|
|
|
|
**Key elements**: Current event or condition, visualization style, data presentation format, visual elements to include.
|
|
|
|
## Professional Best Practices
|
|
|
|
### Be Hyper-Specific
|
|
|
|
Instead of generic descriptors, provide detailed specifications.
|
|
|
|
**Bad**: "fantasy armor"
|
|
**Good**: "ornate elven plate armor, etched with silver leaf patterns, with a high collar and pauldrons shaped like falcon wings"
|
|
|
|
### Provide Context and Intent
|
|
|
|
Explain the image's purpose.
|
|
|
|
**Bad**: "create a logo"
|
|
**Good**: "Create a logo for a high-end, minimalist skincare brand"
|
|
|
|
### Iterate and Refine
|
|
|
|
Leverage the conversational model nature. Make incremental changes with follow-up prompts like "Keep everything the same, but change the expression to be more serious."
|
|
|
|
### Use Step-by-Step Instructions
|
|
|
|
For complex multi-element scenes, break requests into sequential steps: "First, create... Then add... Finally, place..."
|
|
|
|
### Employ Semantic Negative Prompts
|
|
|
|
Describe the desired outcome positively rather than stating what to exclude.
|
|
|
|
**Bad**: "no cars"
|
|
**Good**: "an empty, deserted street with no signs of traffic"
|
|
|
|
### Control the Camera
|
|
|
|
Use cinematic language to command composition: "wide-angle shot," "macro shot," "low-angle perspective"
|
|
|
|
## Prompt Rewriting Checklist
|
|
|
|
When rewriting a user's simple prompt, consider adding:
|
|
|
|
1. **Subject details**: What exactly is being shown? Be specific about appearance, clothing, expression
|
|
2. **Environment/Setting**: Where is this? Indoor/outdoor, time of day, weather, location type
|
|
3. **Lighting**: Natural/artificial, direction, quality (soft/harsh), color temperature
|
|
4. **Composition**: Camera angle, distance, framing, focal length
|
|
5. **Style/Aesthetic**: Photorealistic, illustration style, art movement, medium
|
|
6. **Mood/Atmosphere**: Emotional tone, energy level, color palette mood
|
|
7. **Technical specs**: Aspect ratio, level of detail, texture descriptions
|
|
|
|
## Example Prompt Transformations
|
|
|
|
### Simple to Effective
|
|
|
|
**User prompt**: "a cat"
|
|
**Rewritten**: "A fluffy orange tabby cat lounging on a sun-drenched windowsill, soft afternoon light creating a warm glow on its fur. The cat is in a relaxed pose with half-closed eyes, conveying contentment. Shot from a low angle with shallow depth of field, the background showing a blurred garden view through the window. Photorealistic, warm color palette."
|
|
|
|
**User prompt**: "banner for my app"
|
|
**Rewritten**: "A modern, full-bleed banner for a technology application. Dark gradient background transitioning from deep navy to black. Abstract geometric network visualization with glowing nodes connected by thin lines. Clean sans-serif typography positioned left of center. Professional, tech-forward aesthetic with no visible borders or edges - the design extends seamlessly to all edges."
|
|
|
|
**User prompt**: "mountain landscape"
|
|
**Rewritten**: "A majestic mountain landscape at golden hour, with snow-capped peaks catching the last warm rays of sunlight. Rolling clouds weave between the mountain ridges, creating depth and atmosphere. Foreground shows alpine meadow with wildflowers. Wide-angle shot emphasizing the vast scale. Dramatic but natural lighting with warm highlights and cool shadows. 16:9 aspect ratio."
|
|
|
|
## Image Editing Prompts
|
|
|
|
### Adding/Removing Elements
|
|
Describe changes while maintaining original style, lighting, and perspective details.
|
|
|
|
### Style Transfer
|
|
Provide photographic source and request recreation in specific artistic styles, preserving original composition.
|
|
|
|
### High-Fidelity Detail Preservation
|
|
Describe critical details (faces, logos) in great specificity alongside edit requests.
|
|
|
|
## Common Mistakes to Avoid
|
|
|
|
1. **Keyword lists**: "cat, cute, fluffy, orange" - Use descriptive sentences instead
|
|
2. **Vague requests**: "make it better" - Be specific about what to improve
|
|
3. **Conflicting instructions**: "photorealistic cartoon" - Choose a consistent style
|
|
4. **Missing context**: "logo" - For what? What industry? What feeling?
|
|
5. **Negative-only prompts**: "no blur, no artifacts" - Describe what you want, not what to avoid
|