AI Image Prompts · 绘画提示词库

Create a set of presentation slides, adopting a whiteboard and doodle style, focusing on step-by-step explanations, featuring a consistent colored thick pencil hand-drawn style, rich information, including Chinese text, using the nano_banana tool to generate images, and extracting core information.

- {argument name="use case 1" default="LinkedIn Infographic"}
- {argument name="use case 2" default="Technical Architecture Diagram"}
- {argument name="use case 3" default="Product UI Comparison Model"}
- {argument name="use case 4" default="Data Visualization"}
- {argument name="use case 5" default="Code Visualization Explanation"}

Based on the reference image, generate a multi-cut image with 9 cuts arranged within a single image.
The layout is a 3x3 uniform grid. Each cut features the same person, same composition, same pose, and same camera position.
Specify the style for each cut (from top left to bottom right):
Cel animation style (TV anime style, clean line art, 2-3 levels of shadow)
Movie anime style (delicate lines, rich gradient shadows)
Black and white manga style (ink lines, screen tones)
Watercolor illustration (paper texture, bleeding, pale colors)
Oil painting (thick application, brushstrokes, canvas feel)
High-quality 3D cartoon (soft modeling, subsurface scattering)
Photorealistic 3D (PBR, texture close to real life)
Claymation (clay texture, handmade feel)
Cyberpunk color grading (cyan/magenta rim light, background remains plain)
Ensure that all 9 cuts have completely identical composition, person, pose, and background, allowing only the style difference to be compared.
Arrange them naturally as one complete image.

1. Overall Visual Style

Style Keywords: Flat UI, Vector, Modern Minimalist, Card-style design.

Lines and Shapes:

Rounded Corners Priority: All borders, background blocks, buttons, and process nodes must be rounded rectangles (Rounded Rectangles, with a large corner radius, approximately 10-15px).

Elbow Lines: Process connections use right-angle polylines (no curves), lines are simple and strong.

Color Logic (Key):

“Candy Color” Scheme: Use light backgrounds with high brightness and low saturation (such as light blue, light green, light purple), paired with darker borders of the same color family.

Border Emphasis: All graphics must have clear 2px~3px solid borders.

2. Vertical Layout Structure

Regardless of the topic, please follow the following top-down layout order, which must be contained within a large border container:

Layer 1: Main Title (Header)

Position: Centered at the top of the container.

Style: Bold black font, largest font size, darkest color.

Layer 2: Core Flowchart Area

Position: Occupies 50% of the core area of the layout.

Content: Rectangular nodes + arrow connections. If it is a loop step, please use a dashed line pointing back.

Layer 3: Feature List

Position: Below the flowchart, centered or left-aligned.

Style: Short and concise bullet points, slightly smaller font, with some white space.

Layer 4: Full-width Summary Box (Footer Box)

Position: At the very bottom of the container, filling the width horizontally.

Color Scheme: Use a contrasting color (such as purple in the example), distinct from the main color scheme above.

Function: Used to place 'Summary', 'Analogy', or 'Core Conclusion'.

Landscape size (horizontal/scenic mode) is chosen because:

Content Adaptation: Horizontal content like timelines is better displayed in a horizontal canvas to show the evolution process.

Presentation Friendly: The 3:2 horizontal ratio is very suitable for insertion into PPTs and documents, and displays well on projectors or screens.

Information Capacity: Horizontal layout can accommodate more text and icons, avoiding clutter.

A Persona-style status screen using nano banana pro. I instructed it to randomly generate the level and stats, but they tend to be high.

Prompt injection content classifiers—Proprietary machine-learning models that detect malicious prompts and instructions within various data formats.

Security thought reinforcement—Targeted security instructions that are added around the prompt content. These instructions remind the LLM (large language model) to perform the user-directed task and ignore adversarial instructions.

Markdown sanitization and suspicious URL redaction—Identifying and redacting external image URLs and suspicious links using Google Safe Browsing to prevent URL-based attacks and data exfiltration.

User confirmation framework—A contextual system that requires explicit user confirmation for potentially risky operations, such as deleting calendar events.

End-user security mitigation notifications—Contextual information provided to users when security issues are detected and mitigated. These notifications encourage users to learn more via dedicated help center articles.

Model resilience—The adversarial robustness of Gemini models, which protects them from explicit malicious manipulation.

{
  "scene": "AI engineer working at futuristic desk with multiple holographic displays",
  "camera": {
    "angle": "{argument name="camera angle" default="over-the-shoulder, slightly elevated"}",
    "focal_length": "{argument name="focal length" default="50mm portrait lens"}",
    "depth_of_field": "{argument name="depth of field" default="shallow, subject in sharp focus, background slightly blurred"}"
  },
  "lighting": {
    "type": "{argument name="lighting type" default="dramatic side lighting with blue accent lights"}",
    "time_of_day": "{argument name="time of day" default="late night"}",
    "mood": "{argument name="mood" default="focused, tech-noir atmosphere"}",
    "shadows": "deep but not harsh"
  },
  "color_grade": "{argument name="color grade" default="cool tones, teal and orange color scheme"}"
}

front facing, straight monitor, close shot

A game that automatically analyzes and corrects images.

You are an experienced senior UI/UX designer specializing in new service launches and redesigns.

Analyze the attached image (or input text if no image is attached) and extract **only the 'functional requirements' and 'information structure'** shown there.
Based on the extracted functions, generate a high-resolution device mockup image of the app/service with a **completely refreshed design**, showing both the 'Smartphone version (mobile app)' and the 'PC version (Web dashboard or LP)' side-by-side.

**Important Requests:**
1.  **Design Refresh:** **Do not imitate** the visual style of the attached image at all. Propose a completely new design according to the 'Pop Design' instructions below.
2.  **Functionality Maintenance:** Core app functions, such as 'Upload,' 'Video Tiling Display,' and 'Export Settings,' must be included in the UI.
3.  **No Characters:** Do not place any characters or human illustrations in the image. Focus only on the UI and devices.

▼ Instructions for the Generation Process (Thinking Process)
Construct the design by executing the following thought process before drawing:
 * [Step 1: Extraction of Functional Requirements]: Identify what the service does from the attached image (e.g., video upload, tiling edit, output) and list the necessary UI elements.
 * [Step 2: Formulation of Pop Design Concept]: Define a vibrant color palette, playful shapes, and approachable typography based on the '▼ New Design and Style Specification' below.
 * [Step 3: Responsive UI Design]:
    * Smartphone Version: Optimize the UI using the defined pop design so that it is fun to operate even on a small screen. Buttons should be large and easy to press.
    * PC Version: Design a visually enjoyable grid layout utilizing the wide screen, using the defined pop design.
 * [Step 4: Spatial Composition]: Set an angle where the PC and smartphone are beautifully aligned, and the new design of both screens is clearly conveyed.

▼ New Design and Style Specification
 * **Design Direction:** Adopt **'Fun, Approachable, and Energetic Pop Design.'**
    * **Color:** Use a highly saturated, vibrant multi-color palette (e.g., candy colors, bright yellow, hot pink, sky blue) to give an energetic impression.
    * **Shape:** Frequently use large rounded corners for buttons, cards, and windows to create a soft, tactile texture. Icons should be in a playful style drawn with thick lines.
    * **Font:** Use an approachable and highly legible rounded Gothic Japanese font.
    * **Overall Atmosphere:** Finish with a bright, positive UI that makes the user excited to interact with it.
 * **Composition:** Cross-platform mockup with the PC and smartphone beautifully aligned, in an isometric (oblique overhead) or front view. The background should be a bright studio environment that enhances the pop UI.
 * **Text:** Place accurate Japanese text indicating the functions.

▼ Input Text (App Function Definition)
{argument name="app function definition" default="App to tile and export uploaded videos\nNo AI analysis function needed\nI want to be able to set the resolution!"}

A vibrant rhythm game screenshot featuring a 3D anime character from the reference photo.
[{argument name="character description" default="Girl wearing a cloak"}]

Maintaining a confident dance pose on the left side of the screen.
Immersive cinematic lighting glowing in neon yellow and pink
Glossy, reflective dance floor shining under a spotlight
Dynamic 3D anime style.

Immersive UI rhythm game screen: no reflection display
Score display at the top
Colorful music waveform animation synchronized with the beat
Stage timer countdown
Floating combo numbers
Energy bar
Neon particle effects
Four falling button instruction arrows from top to bottom on the right side of the screen
Button arrow direction and color are fixed for each column (Left: Red, Up: Blue, Down: Green, Right: Yellow)

Highly detailed game atmosphere created with immersive arcade rhythm game HUD elements
Ultra-detailed, cinematic, immersive, 3D animation.

hand-drawn urban planning map