🧠 Overview

Writing a good prompt is one of the most important skills for getting stunning results on MyStage.AI.
Whether you’re using text-to-image, text-to-video, or image-to-video, clear and descriptive prompts help the AI understand exactly what you want to create.

This guide covers best practices, structure, and real examples to help you generate consistently high-quality outputs.

---

✍️ How Prompts Work

When you type a prompt, you’re giving the AI a set of visual and conceptual instructions.
The more specific and structured your prompt, the closer the output will be to your vision.

💡 Think of it like directing a movie, the more clearly you describe the scene, lighting, and emotion, the better the AI performs.

---

🧩 Prompt Structure

Here’s a simple and effective format to follow:

[Subject] + [Action/Emotion] + [Environment/Lighting] + [Style/Camera/Detail]

Example:

“A young woman standing in a city street at night, neon reflections on wet pavement, cinematic lighting, bokeh background, photographed on a Sony A7R IV, 85mm lens, ultra realistic.”

Video Example:

“A futuristic robot walking through a desert at sunset, wind blowing dust, cinematic 16:9 composition, soft camera movement.”

---

💡 Key Elements of a Strong Prompt

Element	Description	Example
Subject	Who or what the scene is about	“An astronaut”
Action / Emotion	What’s happening or being felt	“Looking at the horizon”
Environment	The setting or background	“Inside a space station”
Lighting	Defines tone and mood	“Soft rim light, blue glow”
Style / Camera	Controls composition and realism	“Cinematic 4K render, 50mm lens”

---

⚙️ Using Modifiers & Styles

You can enhance your prompt with style modifiers:

• Cinematic lighting
• Studio portrait
• Volumetric fog
• HDR detail
• Realistic reflections
• Film grain

Combine them wisely for a cohesive look.

💡 Avoid using too many conflicting modifiers — “cinematic” and “cartoon” in the same prompt will confuse the AI.

---

🚫 Common Mistakes to Avoid

Mistake	Problem	Fix
Too vague	“A person smiling”	Add detail — “A confident woman smiling in front of a sunset skyline”
Too long	AI gets confused	Keep under 1,000 characters
Contradictory styles	“Realistic anime painting”	Choose one visual direction
Missing context	“A lion”	Add environment — “A lion resting under golden savanna light”
Overloading details	Too many concepts	Simplify: 2–3 strong visual ideas per prompt

---

🧠 Pro Tips

• Use reference words from real photography or cinematography (e.g., “35mm lens,” “studio lighting,” “soft shadows”).
• Write prompts in English for best model compatibility.
• Save your best-performing prompts in the History tab for reuse.
• Experiment: slight wording changes can lead to drastically different outputs.
• When using Flux or Kling, focus on cinematic storytelling cues (camera motion, tone, perspective).

---

🎨 Sample Prompts by Tool

Tool	Example Prompt
Flux	“Cinematic portrait of an ancient king, detailed facial structure, marble texture, dramatic shadows, 8K render.”
Kling / Veo	“A samurai walking through falling cherry blossoms, slow motion, cinematic light rays, soft focus.”
Stable Diffusion	“A digital painting of a dragon flying over a mountain at sunrise, fantasy realism, intricate details.”
WAN / Hailuo	“A robot awakening in a neon lab, lens flare reflections, cinematic sequence, 4K video.”

---

🧩 When to Use Negative Prompts

If the AI keeps generating unwanted details, use negative prompts to exclude them.

Example:

“A portrait of a woman, cinematic lighting, ultra-realistic — no text, no watermark, no distortion, no blur.”

💡 Negative prompts are especially useful for Stable Diffusion and Flux image models.