Prompt Builder
Build professional prompts using the empirical format for best results. The Prompt Builder helps you construct detailed, effective prompts by combining proven keywords and phrases in an optimal structure.
β¨ Pro Tool: The Prompt Builder is designed to help both beginners and experts create high-quality prompts that consistently produce excellent results.
How It Works
The Prompt Builder uses a category-based system where you select keywords from different groups (Output Type, Subject, Character, Scene, Object, Style, Mood, etc.). These selections are automatically arranged into a well-structured prompt following proven compositional patterns.
The Prompt Builder Interface
The interface is organized into three main sections: Your Prompt (top), Category Filters (middle), and Selection Panels (bottom):
Understanding the Interface
Getting Started - Step by Step
Step 1: Choose an Output Type
Start by selecting what kind of output you want. This sets the technical foundation for your image:
photo - Photorealistic renders
illustration - Hand-drawn or painted look
digital painting - Digital art aesthetic
3D render - Computer-generated 3D look
35mm film still - Cinematic film photography
Example: Click "photo" in the Output Type panel. It appears as a blue chip in Your Prompt area.
Step 2: Add Your Subject
Select what your image is about - the main focus or subject:
For portraits: "warrior woman", "astronaut", "samurai"
For creatures: "dragon", "phoenix", "mechanical beast"
For objects: "crystal staff", "spaceship", "ancient artifact"
For scenes: "mystical forest", "cyberpunk city", "alien planet"
Step 3: Build Your Scene (Optional)
Use the category filters to add more details:
Style: "cinematic", "fantasy art", "concept art"
Lighting: "golden hour", "dramatic lighting", "neon glow"
Camera Angle: "wide shot", "close-up", "aerial view"
Mood: "epic", "mysterious", "serene"
Quality: "highly detailed", "8k", "sharp focus"
Step 4: Review Your Prompt
Look at the assembled prompt in the "Your Prompt" area. The keywords are automatically arranged in an optimal order:
photo, warrior woman, fantasy armor, standing on cliff, sunset lighting, wide shot, epic, highly detailed, 8k
Step 5: Send to Generator
Click "Send to Generator" to open the Image Generator with your prompt pre-filled, or click "Copy" to use it elsewhere.
Keyword Categories Explained
Output Type
Defines the technical rendering style:
Photography: photo, 35mm film still, 70mm film still, HDR photo
Digital Art: digital painting, digital illustration, concept art
3D/CGI: 3D render, unreal engine render, octane render
Traditional: oil painting, watercolor, pencil sketch
Stylized: pixel art, low poly, vector illustration
Subject
The main focus of your image:
Characters: warrior, mage, astronaut, knight, samurai
Creatures: dragon, phoenix, wolf, mythical beast
Objects: sword, spaceship, crystal, ancient tome
Environments: forest, city, temple, alien world
Style
Artistic direction and influence:
Art Movements: art deco, baroque, impressionist, surrealist
Media: cinematic, anime, comic book, sci-fi
Eras: victorian, futuristic, medieval, cyberpunk
Artists: in the style of [artist name]
Lighting
How the scene is illuminated:
Natural: golden hour, sunset, sunrise, overcast, moonlight
Dramatic: rim lighting, dramatic shadows, chiaroscuro
Studio: soft lighting, studio lighting, professional lighting
Special: neon glow, bioluminescent, volumetric lighting
Camera & Composition
Framing and perspective:
Distance: extreme close-up, close-up, medium shot, wide shot
Angle: low angle, high angle, bird's eye view, worm's eye view
Lens: fisheye, wide-angle, telephoto, macro
Composition: rule of thirds, centered, symmetrical
Mood & Atmosphere
Emotional tone:
Epic: epic, heroic, majestic, grand
Dark: ominous, sinister, haunting, mysterious
Peaceful: serene, tranquil, calm, peaceful
Energetic: dynamic, action-packed, intense, explosive
Quality & Detail
Technical specifications:
Resolution: 4k, 8k, high resolution, ultra HD
Detail: highly detailed, intricate details, sharp focus
Quality: masterpiece, award-winning, professional
Rendering: photorealistic, hyper-realistic, trending on artstation
Prompt Building Strategies
The Pyramid Method
Build from general to specific:
Foundation: Output type (e.g., "photo")
Core: Main subject (e.g., "warrior woman")
Context: Setting/environment (e.g., "ancient temple")
Details: Lighting, camera angle (e.g., "golden hour, wide shot")
Polish: Mood and quality (e.g., "epic, highly detailed")
The Recipe Format
Think of your prompt as a recipe with proportions:
Example Recipe:
1 part technical (photo, 3D render)
2 parts subject (dragon, holding crystal orb)
2 parts environment (volcanic landscape, lava flows)
1 part lighting (dramatic fire glow)
1 part mood (epic, menacing)
1 part quality (8k, highly detailed)
The Story Method
Build a narrative through your selections:
Who: cyberpunk hacker
What: standing in rain
Where: neon-lit city street
When: night time
Why/How: mysterious, cinematic
Advanced Techniques
Keyword Alternatives
Click any blue keyword chip in your prompt to see alternative options. This lets you quickly test variations:
Click "wide shot" β see "close-up", "medium shot", "aerial view"
Click "dramatic lighting" β see "soft lighting", "rim lighting", "god rays"
Click "fantasy art" β see "concept art", "digital art", "matte painting"
Combining Categories
Layer keywords from multiple categories for rich, complex prompts:
Output: digital painting
Subject: ancient wizard
Object: glowing staff
Scene: mystical library
Lighting: magical glow
Style: fantasy art
Mood: mysterious
Quality: highly detailed, 4k
Text Mode Editing
Switch to Text Mode to:
Reorder keywords: Rearrange for emphasis
Add custom phrases: Insert specific details not in the keyword list
Fine-tune weights: Add emphasis with parentheses or increase/decrease keyword strength
Remove fillers: Clean up automatic commas and conjunctions
β οΈ Note: When you switch to text mode and make manual edits, you can't switch back to visual mode without losing your changes. Text mode is a one-way transition for that editing session.
Best Practices
Do's
β Start with Output Type: Always select this first to set the foundation
β Be Specific: "warrior woman in golden armor" beats "woman"
β Use Lighting: Lighting keywords dramatically improve results
β Add Quality Tags: "highly detailed", "8k" push the AI to higher quality
β Layer Keywords: Combine multiple categories for rich prompts
β Test Variations: Try the keyword alternative feature
β Save Good Prompts: Copy successful prompts to reuse and modify
Don'ts
β Don't Overload: More than 15-20 keywords becomes counterproductive
β Don't Contradict: Avoid opposing keywords like "day" and "night"
β Don't Skip Subject: Always define what your image is about
β Don't Use Negatives: The builder doesn't support negative prompts
β Don't Repeat: Each keyword should be unique
β Don't Rush: Take time to build and review your prompt
Example Prompts by Genre
Fantasy Portrait
digital painting, elven queen, ornate silver crown, flowing robes, ancient forest throne room, magical ambient lighting, close-up portrait, fantasy art, regal, highly detailed, 8k
Why it works: Clear subject (elven queen), rich environmental details (throne room, forest), specific lighting (magical ambient), appropriate style tags (fantasy art, digital painting).
Sci-Fi Environment
3D render, futuristic city, neon skyscrapers, flying vehicles, rain, night, cyberpunk, wide angle shot, cinematic, atmospheric, octane render, 4k
Why it works: Technical style (3D render, octane render), layered environment (city, skyscrapers, vehicles, rain), clear mood (cyberpunk, atmospheric), proper framing (wide angle).
Character Action Shot
photo, samurai warrior, katana drawn, cherry blossom petals falling, sunset, dramatic backlighting, medium shot, action pose, cinematic, dynamic, 35mm film still, shallow depth of field
Why it works: Photorealistic style (photo, 35mm film), active subject (samurai, action pose), beautiful environmental details (cherry blossoms, sunset), professional framing (medium shot, shallow DOF).
Product/Object Focus
photo, ancient magical tome, glowing runes, ornate leather binding, floating in air, dark background, spotlight, close-up, mysterious, studio lighting, sharp focus, highly detailed
Why it works: Clear object focus (tome), magical details (glowing runes, floating), professional lighting (spotlight, studio), quality tags (sharp focus, highly detailed).
Common Use Cases
Learning Prompt Structure
Perfect for beginners who want to understand how effective prompts are structured. The builder teaches you the optimal order and combinations.
Quick Iteration
Test multiple variations rapidly by swapping keywords using the alternatives feature. Great for exploring different styles or moods for the same concept.
Building Prompt Libraries
Create and save template prompts for different scenarios. Copy them to a note file and reuse with modifications.
Collaboration & Consistency
Teams can use the same keyword selections to maintain consistent visual style across projects.
Troubleshooting
Prompt Feels Generic
Solution: Add more specific keywords. Instead of just "woman", add "warrior woman, golden armor, holding flaming sword".
Too Many Keywords
Solution: If your prompt has more than 20 keywords, it may be too complex. Remove the least important ones, focusing on subject, lighting, and quality.
Results Don't Match Intent
Solution: Check for contradicting keywords (e.g., "photorealistic" + "cartoon"). Use the keyword alternatives feature to try similar but different terms.
Can't Find the Right Keyword
Solution: Switch to Text Mode and type your custom keyword directly. The builder's keywords are suggestions, not limitations.
Tips for Pro Results
Start Simple, Build Up: Begin with 5-7 keywords, test, then add more if needed
Master One Category: Become expert in one area (e.g., lighting) to dramatically improve all your prompts
Study Examples: Look at prompts from Community Gallery and reverse-engineer them in the builder
Keep a Swipe File: Save successful prompts and keyword combinations
Test Systematically: Change one variable at a time to learn what each keyword does
Use Lighting Liberally: Lighting keywords have the biggest impact on mood and quality
Balance Specificity: Too vague = generic, too specific = limiting. Find the middle ground
End with Quality: Always finish prompts with quality tags like "highly detailed, 8k"
π¨ Pro Workflow: Select output type β Add main subject β Layer environment/scene β Add lighting β Set mood β Add quality tags β Review β Send to Generator β Adjust and iterate. This methodical approach produces consistently great results!
