Prompt Cheatsheet
Quick reference for common quality tags, styles, lighting, and depth-of-field — plus a list of common prompt mistakes.
This cheatsheet pulls together the most useful English tags for AI image generation — quality tags, styles, lighting, depth of field, and common mistakes. We recommend reading Prompt Basics first for syntax, then Prompt Advanced for composition thinking. Use this page as your tag dictionary while you write.
Quality Tags
| Tag | Effect |
|---|---|
masterpiece | Masterpiece-tier quality |
best quality | Top quality |
highly detailed | High level of detail |
sharp focus | Sharp focus |
intricate details | Refined, intricate details |
Style Tags
| Style | Tags |
|---|---|
| Japanese anime | anime, anime style, anime coloring |
| Cel shading | cel shading, flat color |
| Impasto | impasto, thick paint, painterly |
| Watercolor | watercolor, soft edges, color bleed |
| Photorealistic | photorealistic, photo, realistic |
| Cinematic | cinematic, dramatic lighting, film grain |
Lighting Types
Lighting is one of the most powerful tools for setting mood. Swap the lighting and the entire feel of the image changes.
| Lighting type | Tags | Effect and use cases |
|---|---|---|
| Natural light | natural light | Mimics real-world ambient light, feels true-to-life. Good for everyday, realistic, slice-of-life |
| Soft light | soft light, soft lighting | Light is diffused; smooth value transitions, delicate skin. Good for tender, healing, everyday scenes |
| Backlight | back light, backlit | Light comes from behind, hair edges glow. Good for emotional, intimate scenes |
| Side backlight | side back light | A backlight variant — one side bright, one side dark, with strong contour separation |
| Cinematic light | cinematic light, cinematic lighting | Designed lighting with clear bright/dark zones. Good for narrative, premium feel |
| Dappled light | dappled light, dappled sunlight | Light fragmented by an obstruction, like leaves filtering sunlight. Good for outdoors, forests, poetic moods |
| Rim light | rim light, rim lighting | A glow along the character's edge that separates them from the background. Good for night scenes, emphasizing the subject |
| Rembrandt light | rembrandt light, rembrandt lighting | Light from the upper side creates a triangle of light on the cheek. Good for serious, dramatic moods |
| Top light | top light, overhead lighting | Light from directly above hits the forehead and nose, with shadows in the eye sockets. Good for mysterious, dramatic moods |
| Spotlight | spot light, spotlight | Concentrated beam lighting a small area, stage-like. Good for highlighting a character or specific detail |
Light Color
| Tone | Mood | Common pairings |
|---|---|---|
| Warm (yellow/orange) | Companionship, tenderness, memory | warm lighting, golden hour, sunset light |
| Cool (blue/purple) | Solitude, quiet, distance | cool lighting, blue light, moonlight |
When warm and cool appear together, one has to lead! The most common approach is one as the key light and the other only in shadows or as a fill.
Depth of Field and Focus
Depth of field controls what's sharp and what's blurred — it's the key tool for guiding the viewer's attention.
| DoF type | Tags | Effect and use cases |
|---|---|---|
| Bokeh | bokeh, soft bokeh | Background blurred, point lights become circles, dreamy feel. Needs actual highlights to pop |
| Shallow DoF | shallow depth of field | Only the subject is sharp, the rest blurred. The most common portrait DoF |
| Deep focus | deep focus | Foreground to background all sharp, lots of detail. Good for showing a full environment |
| Soft focus | soft focus | Slight blur across the whole image (not just the background), creating a hazy feel. Good for dream-like, memory-tinged scenes |
| Motion blur | motion blur | Directional smearing on moving subjects, conveying speed and motion |
What Depth of Field Does
- Guides attention — Sharp areas naturally draw the eye; blurred areas get ignored
- Controls information density — Deep focus adds information, shallow DoF removes it. Subtraction makes the image cleaner
- Builds spatial layers — Blurred foreground, sharp midground, blurred background creates depth
- Reinforces emotion — Shallow DoF + soft focus feels dreamy and tender; deep focus feels calmer, more objective
Looking Up Tags
Don't know what an English tag is?
Try searching on Danbooru for the relevant tags.
Common Mistakes
| Mistake | Notes |
|---|---|
| Description too vague | "A pretty girl" works less well than 1girl, long pink hair, blue eyes, smile |
| Too many tags | Past 20 tags they start interfering with each other; trimming wins |
| Forgetting quality tags | Without quality tags, output quality drops noticeably |
| Writing in non-English | Most models understand English far better than other languages |
| Conflicting tags | E.g. full-body outfit description + face close-up = distorted output |
FAQ
How many tags should a prompt have?
8–15 tags is usually about right. Past 20 they interfere with each other and quality drops. Start with the subject (character, scene), then add a few quality tags, style, and lighting — trimming beats stacking.
Can I write prompts in other languages?
Most models understand English far better than other languages, so we recommend English tags. Use a translator for words you're not sure about, or check Danbooru for the matching English tags.
Are quality tags required?
Strongly recommended. Tags like masterpiece, best quality, highly detailed noticeably boost detail. Most PixAI models already include default quality and negative prompts, so usually "subject + a few quality tags" is enough.
What if I can't find the English tag I want?
Beyond the Danbooru tag library, you can also use PixAI's Prompt Helper to translate, expand, and refine your prompts automatically.
Why does the image come out distorted or with messy composition?
Usually because of conflicting tags — for instance, "full-body outfit detail" together with "face close-up." When writing prompts, decide on the shot first (close-up / half-body / full-body), then describe details that fit it. For the full composition workflow, see Prompt Advanced.
Was this page helpful?
🎁 Adopted valid suggestions earn a 5,000 Credit reward!
Prompt Advanced
Advanced prompt techniques — composition thinking, layering control, lighting and depth-of-field description for more polished AI art.
Model Overview
Compare all PixAI models — Tsubaki.2, Haruka v2, and Reference Pro — at a glance, including their differences, supported parameters, and available tools.